Top 10 Generative AI Companies in 2026: Products, Pricing, Funding, and Competitive Analysis

The generative Artificial Intelligence (AI) has proven to be one of the most impactful technology segments reshaping content creation, software development, enterprise automation, search, and customer engagement. Guided by the rapid development of large language models (LLMs), multimodal AI, and autonomous AI agents, the industry is witnessing fierce competition between top providers in 2026.

This article analyzes the top 10 generative AI companies in the world, including OpenAI, Anthropic, Google DeepMind, Microsoft Corporation (MSFT), DeepSeek, Mistral AI, Perplexity AI, xAI Inc. (x.ai), Meta AI, and Sarvam AI, and examines their product portfolio, pricing data, funding history, and overall impact on the global AI landscape.

1. OpenAI

Company Overview

Parameter	Details
Headquarters	San Francisco, California, USA
CEO	Sam Altman
Founded	2015
Website	https://openai.com/
Core Focus	Frontier flagship models, multimodal AI, AI tools, and service tiers, and enterprise AI offerings

OpenAI is the largest commercial generative AI company globally and one of the most influential frontier foundation model generators. The company runs its operations, such as ChatGPT for writers, the GPT-series models, enterprise APIs, agentic AI products, and some strategic partnerships in the cloud.

Subscription Plans Pricing

Plan	Price
Free	$0
ChatGPT Go	$8 USD/month (differs in regional markets)
ChatGPT Plus	$20 USD/month
ChatGPT Pro	$200 USD/month
Enterprise	Custom Pricing

OpenAI API Pricing for Flagship Models

Model	Processing Model	Input (per 1M tokens)	Cached Input (per 1M tokens)	Output (per 1M tokens)	Application
GPT-5.6 Sol	Standard	$5.00	$0.50	$30.00	Flagship model for complex coding, professional work, computer use, and long-running agents
	Batch-50%	$2.50	$0.25	$15.00
	Data Residency +10%	$5.50	$0.55	$33.00
GPT-5.6 Terra	Standard	$2.50	$0.25	$15.00	Lower-cost model for coding, professional work, and agentic applications
	Batch-50%	$1.25	$0.125	$7.50
	Data Residency +10%	$2.75	$0.275	$16.50
GPT-5.6 Luna	Standard	$1.00	$0.10	$6.00	Fast and affordable model for high-volume work, coding, and subagents
	Batch-50%	$0.50	$0.05	$3.00
	Data Residency +10%	$1.10	$0.11	$6.60

Products Offering

Product	Description
ChatGPT (Free)	The free version provides limited access to current ChatGPT models and tools, including voice, file uploads, and task assistance, subject to usage limits.
ChatGPT Plus	Add-ons that allow ChatGPT to focus on enhancing and providing new features with faster response time and the availability even during peak hours
GPT API	An industry-leading developer platform for developing multimodal AI to make real-world use easier.
Sora	Generate up to one-minute-long video content by employing a text-to-video AI model that stays true to users' prompts and aesthetics.

In June 2026, OpenAI’s frontier models and Codex are reported to be broadly available through Amazon Bedrock, empowering enterprises to leverage advanced AI in their production environments while maintaining existing security, compliance, and governance workflows. This eases adoption for organizations moving to AI through building/debugging/modernizing code natively in AWS environments.

In addition, OpenAI rolled out Active Sessions, a security feature allowing users to check all sessions related to their account, in June 2026. Further, in May 2026, OpenAI updated GPT-5.5 Instant on ChatGPT and in the API to enhance response mannerism and quality for smoother flow and more natural interaction.

In March 2026, OpenAI reported that it had raised $122bn in funding at a post-money valuation of $852bn. The investment is intended to support frontier-model development, computing infrastructure, and closer integration across ChatGPT, Codex, browsing, and agentic workflows.

In July 2026, OpenAI introduced the GPT-5.6 family, comprising Sol, Terra, and Luna. The company also introduced ChatGPT Work for long-running task execution and GPT-Live for more natural full-duplex voice interactions.

2. Anthropic

Company Overview

Parameter	Details
Headquarters	San Francisco, California, USA
CEO	Dario Amodei
Founded	2021
Website	https://www.anthropic.com/
Core Focus	Safe AI systems, enterprise reasoning, coding models

Anthropic has become OpenAI's most solid direct competition in enterprise AI. Anthropic develops the Claude family of models, famous mainly for their emphasis on development that is safety-centered and well-suited for long-context reasoning, coding performance, and enterprise-grade deployments. Claude is popular in software engineering, knowledge work, and agentic AI workflows.

Subscription Plans Pricing

Plan	Price
Claude (Free)	$0
Claude Pro	$17/month
Claude Max	Max 5x: $100 Per person billed monthly
Claude Max	Max 20x: $200 Per person billed monthly
Team	Standard seat: $20 Per seat/month (if billed annually) $25 (if billed monthly)
Team	Premium seat: $100 Per seat/month (if billed annually) $125 (if billed monthly)
Enterprise	Custom Pricing

Claude API Pricing for Latest Models

Model	Processing Model	Input (per 1M tokens)	Output (per 1M tokens)	Prompt caching (per 1M tokens)		Application
Model	Processing Model	Input (per 1M tokens)	Output (per 1M tokens)	Write	Read	Application
Opus 4.8	Standard	$5	$25	$6.25	$0.50	Most intelligent model for agents and coding
Opus 4.8	Batch-50%	$2.50	$12.50	$3.13	$0.25	Most intelligent model for agents and coding
Sonnet 5	Standard	$3	$15	$3.75	$0.30	Balanced model for coding, agents, and professional work. Introductory API pricing is $2 input and $10 output per 1M tokens through August 31, 2026; standard pricing is $3 and $15 thereafter.
Sonnet 5	Batch-50%	$1.50	$7.50	$1.88	$0.15
Haiku 4.5	Standard	$1	$5	$1.25	$0.10	Strongest mini model for coding & subagents
Haiku 4.5	Batch-50%	$0.50	$2.50	$0.63	$0.05	Strongest mini model for coding & subagents

Products Offering

Product	Description
Claude	The main chat interface to have conversations, write, analyse, research and create with Claude.
Claude Code	An agentic coding system that independently reads codebase, modifies multiple files, executes tests and serves committed code.
Claude Cowork	Desktop automation for non-coders automating workflows around files & tasks on their computer.
Claude Security	AI-powered security tooling that helps teams discover, investigate and remediate threats with speed.

On 9 June 2026, Anthropic announced the launch of its Claude Fable 5 and Claude Mythos 5 which are its New Mythos-Class AI Models. Claude Fable 5 is the widely available version with extra safety mitigations against risks including cyber and biological attacks, whereas Claude Mythos 5 provides advanced AI capabilities and is to be offered through much tighter access routes, with offering only to approved organizations through Project Glasswing.

Both models share the same foundation as their underlying technology, and cost $10 per million input tokens and $50 per million output tokens, which represent less than half of amount that the company charged for its earlier Mythos Preview offering. The company stated the release is focused on offering powerful AI capabilities while implementing its safe rollout strategy.

Access to Fable 5 and Mythos 5 was suspended on June 12, 2026, following a United States government directive. The restrictions were lifted at the end of June, and Fable 5 was restored globally on July 1, while Mythos 5 returned for approved organizations.

Claude Opus 4.8, launched in May 2026, remains Anthropic's premium Opus-class model for complex coding and enterprise agentic work. Fable 5 now sits above the Opus class as Anthropic's most capable generally available model. Anthropic also launched Claude Sonnet 5 on June 30, 2026, as its broader cost-performance model for coding, agents, and professional work.

On May 28, 2026, Anthropic announced a $65 billion Series H funding round at a $965 billion post-money valuation. The funding is intended to expand research, computing infrastructure, and enterprise adoption of Claude.

On May 6, 2026, Anthropic announced a compute partnership with SpaceXAI that gives Anthropic access to the full capacity of the Colossus 1 data center, representing more than 300 MW and over 220,000 NVIDIA GPUs. The agreement is focused on infrastructure capacity for Claude.

SAP and Anthropic partnered to bring Claude into SAP Business AI in May 2026. It aspires to equip SAP with cutting-edge natural language capabilities over its enterprise platform, empowering organizations in areas like workflow automation, data analysis, and AI-driven decision-making.

3. Google DeepMind

Company Overview

Parameter	Details
Headquarters	London, United Kingdom
CEO	Demis Hassabis
Founded	2010
Website	https://deepmind.google
Core Focus	Foundation Models, Multimodal AI, Enterprise AI, AI Agents

Google DeepMind is basically Google’s flagship AI research and commercialization segment, and it’s also the team behind the Gemini model family. Through tighter integration across Search, Android, Workspace, Cloud, YouTube, and the developer ecosystem, Gemini has ended up being one of the most widely spread generative AI platforms around the world. Lately, they’ve been putting more emphasis on multimodal AI, agentic workflows, video generation, and enterprise AI infrastructure.

Subscription Plans Pricing

Plan	Price
Google AI (Free)	$0
Google AI Plus	$7.99/month
Google AI Pro	$19.99/month
Google AI Ultra	$99.99/month
Enterprise	Custom Pricing

Gemini API Pricing

Model	Processing Model	Input (per 1M tokens)	Context caching price (per 1M tokens)	Output (per 1M tokens)	Application
Gemini 3.5 Flash	Standard	$1.50	$0.15	$9.00	It is an intelligent model is built for speed though, mixing frontier intelligence with sharper search and grounding.
	Batch	$0.75	$0.075	$4.50
	Flex	$0.75	$0.08	$4.50
	Priority	$2.70	$0.27	$16.20
Gemini 3.1 Flash-Lite	Standard	$0.25 (text/image/video) $0.50 (audio)	$0.025 (text/image/video) $0.05 (audio)	$1.50	Cost-efficient model for high-volume multimodal understanding and application workloads.
	Batch	$0.125 (text/image/video) $0.25 (audio)	$0.0125 (text/image/video) $0.025 (audio)	$0.75
	Flex	$0.125 (text/image/video) $0.25 (audio)	$0.0125 (text/image/video) $0.025 (audio)	$0.75
	Priority	$0.45 (text/image/video) $0.90 (audio)	$0.045 (text/image/video) $0.09 (audio)	$2.70
Gemini 3.1 Pro Preview	Standard	$2.00, prompts <= 200k tokens $4.00, prompts > 200k tokens	$0.20, prompts <= 200k tokens $0.40, prompts > 200k	$12.00, prompts <= 200k tokens $18.00, prompts > 200k	It is an advanced reasoning and coding model for complex multimodal and agentic work. Google’s low-latency audio-to-audio capabilities are provided through its Gemini Live and Gemini 3.5 Live Translate model families.
	Batch	$1.00, prompts <= 200k tokens $2.00, prompts > 200k tokens	$0.20, prompts <= 200k tokens $0.40, prompts > 200k	$6.00, prompts <= 200k tokens $9.00, prompts > 200k
	Flex	$1.00, prompts <= 200k tokens $2.00, prompts > 200k tokens	$0.20, prompts <= 200k tokens $0.40, prompts > 200k	$6.00, prompts <= 200k tokens $9.00, prompts > 200k
	Priority	$3.60, prompts <= 200k tokens $7.20, prompts > 200k tokens	$0.36, prompts <= 200k tokens $0.72, prompts > 200k	$21.60, prompts <= 200k tokens $32.40, prompts > 200k

Products Offering

Product	Description
Veo	It can generate cinematic-quality video with audio.
Gemini Omni	It is a multimodal model that can create basically anything from any input, starting with video.
Imagen	It generates and edits high-quality images from text prompts. Google’s Lyria family is used for music and audio generation.
Lyria	It can produce high-fidelity music and audio.

In May 2026, Google DeepMind announced the opening of a new AI research lab in Singapore, meant to push AI further across the Asia-Pacific region. The lab will focus on developing frontier AI responsibly, working with local institutions, and supporting innovation in healthcare, sustainability, and education, among other areas.

In May 2026, Gemini Spark launched as a proactive AI agent that can handle tasks for users. Meanwhile, in the same period, Gemini Omni rolled out conversational video creation with AI avatars, along with multimodal editing.

In June 2026, Google added computer-use capabilities to Gemini 3.5 Flash and introduced Gemini 3.5 Live Translate for low-latency voice translation. These developments extend Gemini from content generation and reasoning into interface control and real-time audio workflows.

4. Microsoft

Company Overview

Parameter	Details
Headquarters	Redmond, Washington, USA
CEO	Satya Nadella
Website	https://copilot.microsoft.com/
Founded	1975
Focus	Enterprise AI, Productivity AI, Coding AI

Microsoft created one of the biggest enterprise AI ecosystems worldwide through Copilot, Azure AI, GitHub Copilot, and Microsoft 365 integrations. Additionally, it uses both OpenAI technologies as well as internally developed MAI models to reinforce its whole AI platform strategy.

Subscription Plans Pricing

Plan	Price
Microsoft 365 (free)	$0
Microsoft 365 Basic	$1.99/month
Microsoft 365 Personal	$9.99/month
Microsoft 365 Family	$12.99/month
Microsoft 365 Premium	$19.99/month

Azure OpenAI API Pricing

Model	Processing Model	Input (per 1M tokens)	Cached Input (per 1M tokens)	Output (per 1M tokens)	Application
GPT 5.5 Series	Standard	$5	$0.50	$30	Provides the user with advanced reasoning, instruction following, and agentic capabilities for production AI workloads.
GPT 5.5 Series	Priority Processing	$12.50	$1.25	$75
GPT 5.4 Series	Standard	$2.50	$0.25	$15	Offers stronger reasoning, dependable execution, and agentic workflows at scale.
	Priority Processing	$5	$0.50	$30
	Pricing with Batch API	$1.25	$0.13	$7.50
GPT-5.3 Series	Standard	$1.75	$0.18	$14	It unifies the frontier coding performance of GPT-5.2-Codex with the reasoning and professional knowledge abilities of GPT-5.2.
GPT-5.3 Series	Priority Processing	$3.50	$0.35	$28
GPT-5.2	Standard	$1.75	$0.18	$14	Offers the deep reasoning and expanded context handling needed for building more sophisticated AI agents
GPT-5.2	Priority Processing	$3.50	$0.35	$28
GPT-5.1	Standard	$1.25	$0.13	$10	Provide faster response to users in diverse situations, faster to users with adaptive reasoning.
GPT-5.1	Priority Processing	$2.50	$0.25	$20

Products Offering

Product	Description
Microsoft 365 Copilot	It is an AI assistant that is designed to work in integration with Microsoft 365 apps.
Microsoft Copilot Studio	It is a platform that is utilized to build, customize, and deploy users' own AI agents.
Microsoft Azure	It is an enterprise-grade platform that is developed for evaluating, building, and governing AI applications.

At Microsoft Build on June 2, 2026, Microsoft detailed seven internally developed MAI models: MAI-Thinking-1, MAI-Image-2.5 and its Flash variant, MAI Transcribe 1.5, MAI-Voice-2 and its Flash variant, and MAI-Code-1. The models cover reasoning, image generation and editing, transcription, voice, and coding workloads.

On July 2, 2026, Microsoft announced Microsoft Frontier Company, supported by a $2.5 billion investment and approximately 6,000 industry and engineering specialists to help enterprises design and deploy AI systems.

In May 2026, Microsoft introduced Microsoft 365 Business with Copilot, which is a new subscription built specifically for small businesses. It blends productivity apps, AI-powered Copilot features, and baked-in security to help small teams streamline operations, manage finances, and rise more efficiently.

In April 2026, Microsoft announced plans to invest $10 billion in Japan for the strengthening of AI infrastructure, cybersecurity, and workforce development. This initiative also deepens Microsoft’s long-term commitment to the country, supporting innovation and resilience throughout the Asia-Pacific region.

5. DeepSeek

Company Overview

Parameter	Details
Headquarters	Hangzhou, China
CEO	Liang Wenfeng
Website	https://www.deepseek.com/
Founded	2023
Focus	Open-weight frontier reasoning for large language models.

DeepSeek is one of the more disruptive AI companies globally, after it released reasoning models that seemed really capable, with inference costs that were way lower than a lot of Western competitors. Further, the open weight strategy kind of shifted the market pricing and adoption paths, with more users adopting it.

The company has various models tailored for diverse applications like deepseek-v4-flash, deepseek-v4-pro, deepseek-chat, and deepseek-reasoner, with plans to deprecate older versions, i.e., deepseek-chat and deepseek-reasoner, after 24th July 2026. DeepSeek offers free consumer chat access, while API usage is paid according to the input and output token rates shown below.

DeepSeek API Pricing

Model	1M Input Tokens (Cache Hit)	1M Input Tokens (Cache Miss)	1 M Output Tokens	Concurrency Limit
DeepSeek-V4-Flash	$0.0028	$0.14	$0.28	2500
DeepSeek-V4-Pro	$0.003625	$0.435	$0.87	500
DeepSeek-V3.1 API	$0.07	$0.56	$1.68	-

Products Offering

Product	Description
DeepSeek-V3.1	It is a hybrid model with stronger agent skills, tool use, and 128K context.
DeepSeek-R1 0528	It is an upgraded AI mode with front-end capacities and fewer hallucinations, JSON output, and function calling support.
DeepSeek V3-0324	It is a V3 update with major reasoning, front-end, and tool-use improvements; released under the MIT License.

In April 2026, DeepSeek introduced the V4 Preview and introduced about 1M context length being treated as the norm across basically all its services. They showed two releases: V4-Pro (1.6T total / 49B active params), which was pitched for strong reasoning plus coding benchmarks, while V4-Flash (284B total / 13B active params) was tuned for speed, and for being cheap to run, practical efficiency. Both models also do dual modes (Thinking / Non-Thinking), and they’re tied into major AI agent tools.

Additionally, in December 2025, DeepSeek also introduced V3.2 and V3.2-Special, which are reasoning-first models aimed at agentic tasks. V3.2 was described as giving balanced inference at around the GPT-5 level, while V3.2-Speciale took the reasoning majorly, trying to match the level of Gemini-3.0-Pro, and it reportedly landed gold rank results in competitions like IMO, CMO, and ICPC.

6. Mistral AI

Company Overview

Parameter	Details
Headquarters	Paris, France
CEO	Arthur Mensch
Website	https://mistral.ai/
Founded	2023
Core Focus	Frontier AI LLMs, assistants, agents, services

Mistral AI is Europe’s leading frontier AI startup, and it’s a strong supporter of open-weight AI development. The company targets both developers and enterprises, and it keeps promoting European AI sovereignty through offerings such as enterprise-grade and self-hostable foundation models.

Subscription Plans Pricing

Plan	Price
Mistral AI (free)	$0
Mistral Pro	$14.99/month
Mistral Team	$24.99/month
Mistral Enterprise	Custom pricing
Mistral Education Plan	$5.99/month

Mistral AI API Pricing

Model	Input (per 1M tokens)	Output (per 1M tokens)	Detail
Mistral Medium 3.5	$1.5	$7.5	It merges instruction-following, reasoning, and coding into a single 128B dense model.
Mistral Large 3	$0.5	$1.5	Open-weight, general-purpose, flagship multimodal and multilingual model.
Devstral 2	$0.4	$2	Open-weights agentic coding model for autonomous software engineering.
Devstral Small 2	$0.1	$0.3	It is a lightweight, open model for coding agents.
Codestral	API- $0.3	$0.9	Low-latency coding model optimized for high-frequency completion, fill-in-the-middle, and code generation tasks.
Codestral	Fine-tuning -$0.2	$0.6
Mistral Small 4	$0.15	$0.60	SOTA. Multimodal. Multilingual. Apache 2.0.
Mistral Small Creative	$0.1	$0.3	A fine-tuned small model for creative writing, roleplay, and chat—trained on curated data
Magistral Medium	$2	$5	Thinking model excelling in domain-specific, transparent, and multilingual reasoning
Magistral Small	$0.5	$1.5	It is a text-to-text, reasoning AI, and a multimodal model.
Ministral 3 - 3B	$0.1	$0.1	It is a frontier text-to-text and generative AI model
Ministral 3 - 8B	$0.15	$0.15	It is a frontier text-to-text and generative AI model
Ministral 3 - 14B	$0.2	$0.2	Text-to-text and agentic AI model
Voxtral Small	$0.004 (audio) /$0.1 (text)	$0.4	It is a text-to-text model which offer transcription of speech and audio
Voxtral Mini	$0.001 (audio) / $0.04 (text)	$0.04	It is a voice and text-to-text AI model
Classifier API model 3B	$0.1	$0.1	Fine-tune Ministral 3B for classification tasks, like moderation, sentiment analysis, and fraud detection, among others.
Classifier API model 8B	$0.04	$0.04	Fine-tune Ministral 8B for classification tasks, like moderation, sentiment analysis, fraud detection, and more.

Products Offering

Product	Description
Codestral	It is the first-ever code model and is an open-weight generative AI model for code generation work.
Mistral Large	It is a flagship and cutting-edge text generation model utilized for complex multilingual reasoning tasks.
Mistral Large 3	It is the next generation of the Mistral model that empowers the developer community.

In May 2026, Mistral introduced Vibe Remote Agents, a framework meant to deploy AI agents across distributed setups. At the same time, they released Mistral Medium 3.5, which is basically an upgraded model, with better reasoning, efficiency, and more scalability for enterprise use cases.

In May 2026, Mistral launched Physics AI, a new initiative to combine AI with research and simulation in physics. The ambition of the project is to turbocharge science discovery via advanced machine learning with domain-specific modeling to make previously unattainable strides in materials, energy, and fundamental physics.

During this period, Mistral announced an agreement to acquire Emmi AI, a Physics AI company focused on modelling physical systems, real-time simulation, and industrial engineering. The acquisition expands Mistral beyond conversational AI and strengthens its AI solutions for manufacturing, engineering, simulation, and other industrial applications.

After the original article was published, Mistral introduced OCR 4 on June 23, expanded connector controls on June 24, released Leanstral 1.5 on July 2, and introduced Robostral Navigate on July 8, 2026.

7. Perplexity AI

Company Overview

Parameter	Details
Headquarters	San Francisco, California, USA
CEO	Aravind Srinivas
Founded	2022
Website	https://www.perplexity.ai/
Core Focus	Frontier AI, AI agents

Perplexity AI is considered one of the fastest-growing search and answer engine company with a native, unified experience in AI. The response positions the company as a direct competitor with traditional search engines around citation-based answers, research workflows, and AI-assisted information discovery.

Subscription Plans Pricing

Plan	Price
Perplexity AI (free)	$0
Perplexity Pro	$17/month (if billed annually)
Perplexity Enterprise Pro	$40/month per seat or $400/year
Perplexity Enterprise Max	$325/month per seat or $3,250/year

Perplexity AI Token Pricing

Model	Input (per 1M tokens)	Output (per 1M tokens)
Sonar	$1	$1
Sonar Pro	$3	$15
Sonar Reasoning Pro	$2	$8
Sonar Deep Research	$2	$8

Products Offering

Product	Description
Perplexity Search	AI answer engine providing source-grounded responses and web research.
Perplexity Computer	General-purpose task system that coordinates models, browsers, files, and connected tools.
Enterprise Search	Business-focused search and research environment using internal and external information.

In February 2026, Perplexity introduced Perplexity Computer, which could orchestrate work across 19 models at launch, break complex projects into subtasks, and use browsers, files, and connected services. It was initially available to Max subscribers, with Pro and Enterprise access announced for subsequent rollout. In June 2026, Perplexity added Deep Research, a command panel, forking, inline actions, analytics APIs, and enterprise credit controls.

8. xAI

Company Overview

Parameter	Details
Headquarters	Palo Alto, California
CEO	Elon Musk
Website	https://x.ai/
Founded	2023
Focus	Foundation Models, Generative AI, Conversational AI, AI Infrastructure

xAI joined SpaceX in February 2026, and its official product site now uses the SpaceXAI brand. The company develops large language models under the Grok family, integrated with X (formerly Twitter), enterprise APIs, and AI infrastructure powered by its Colossus supercomputer. xAI focuses on reasoning, real-time information retrieval, multimodal AI, and large-scale AI training capabilities.

Subscription Plans Pricing

Plan	Price
Grok Free	$0
X Premium	$8/month
X Premium+	$40/month
SuperGrok	$30/month
SuperGrok Annual	$300/year

xAI API Pricing

Model	Processing Model	Input (per 1M tokens)	Output (per 1M tokens)	Application
Chat API	grok-4.5	$2.00	$6.00	Chat and agent models for reasoning, coding, grounding, and real-time information retrieval.
	grok-4.3	$1.25	$2.50
	grok-4.20-multi-agent-0309	$1.25	$2.50
	grok-4.20-0309-reasoning	$1.25	$2.50
Imagine API	grok-imagine-image	$0.002 / img	$0.02/img	Advanced flagship model designed for complex reasoning, coding, enterprise workflows, and agentic AI applications.
	grok-imagine-image-quality	$0.01 / img	$0.05/ img
	grok-imagine-video	$0.01 / sec $0.002 / img	$0.05/ img
	grok-imagine-video-1.5-preview	$0.01 / img	$0.08/ mg

Products Offering

Product	Description
Grok	A conversational AI assistant designed for reasoning, real-time knowledge access, coding assistance, and creative tasks.
Grok API	A developer platform that enables businesses to integrate Grok models into applications, workflows, and AI-powered products.
Grok for X	AI assistant integrated within X, providing content analysis, search, summarization, and conversational capabilities.

In June 2026, xAI announced that Grok Voice has been integrated into Vapi, a leading platform for enterprise voice AI applications. According to a blind evaluation conducted by Vapi, Grok Voice achieved the highest ranking among competing voice providers, demonstrating superior voice quality, conversational naturalness, and performance.

In July 2026, SpaceXAI launched Voice Agent Builder and expanded its voice portfolio. On July 8, it introduced Grok 4.5 at $2 per million input tokens and $6 per million output tokens for coding, agentic tasks, and professional work.

On May 6, 2026, SpaceXAI signed a compute-infrastructure agreement that gives Anthropic access to Colossus 1. The agreement is focused on computing capacity for Anthropic’s Claude services.

9. Meta AI

Company Overview

Parameter	Details
Headquarters	Menlo Park, California, United States
CEO	Mark Zuckerberg
Website	https://www.meta.ai/
AI research lineage	FAIR established in 2013
Focus	Open-weight AI Models, Generative AI, Enterprise AI, Consumer AI Assistants

Meta AI is the artificial intelligence division of Meta Platforms, responsible for developing the Llama family of foundation models, Meta AI assistant, AI Studio, and enterprise AI solutions. The company focuses on open-weight AI models, multimodal intelligence, personalized AI experiences, and large-scale AI infrastructure integrated across Facebook, Instagram, WhatsApp, Messenger, and smart devices.

Subscription Plans Pricing

Plan	Price
Meta AI	Free
Meta Model API (public preview)	Pricing not announced
Meta Verified (not a dedicated AI plan)	Starting at $11.99/month
AI Studio	Free
Meta Business AI Solutions	Custom Enterprise Pricing

Products Offering

Product	Description
Meta AI	AI assistant integrated across WhatsApp, Instagram, Facebook, Messenger, and standalone applications.
Llama Models	Open-weight large language models designed for developers, enterprises, and researchers.
AI Studio	A platform that allows creators and businesses to build customized AI assistants and digital agents.
Muse Spark 1.1	Multimodal reasoning model for agentic tasks, coding, computer use, and long-context workflows.

In April 2026, Meta introduced Muse Spark, the first foundation model developed by its newly established Meta Superintelligence Labs (MSL). Designed as a natively multimodal AI system, Muse Spark combines advanced reasoning, tool use, visual chain-of-thought processing, and multi-agent orchestration capabilities.

Meta introduced Muse Image and previewed Muse Video on July 7, 2026. On July 9, it released Muse Spark 1.1, a multimodal reasoning model for agents, coding, computer use, and multimodal understanding, with a one-million-token context window. Muse Spark 1.1 is available in Meta AI’s Thinking mode and through the Meta Model API in public preview.

10. Sarvam AI

Company Overview

Parameter	Details
Headquarters	Bengaluru, Karnataka, India
Leadership	Vivek Raghavan and Pratyush Kumar, Co-founders
Website	https://www.sarvam.ai/
Founded	2023
Focus	Generative AI, Large Language Models, Indic AI, Voice AI

Sarvam AI is an India-based artificial intelligence company focused on building foundational AI models and platforms tailored for Indian languages and enterprises. The company develops multilingual large language models, speech technologies, and enterprise AI solutions that support India's linguistic diversity. Sarvam AI aims to create sovereign AI infrastructure and foundation models optimized for Indic languages, government applications, and enterprise deployment.

Subscription Plans Pricing

Plan	Price
Sarvam API Access	Usage-Based
Enterprise AI Solutions	Custom Pricing
Voice AI Solutions	Rs 15 per 10,000 characters

Sarvam AI API Pricing

API Service	Pricing	Unit	Application
Sarvam-105B	Rs 4 / Rs 2.5 / Rs 16	Per 1M input / cached input/output tokens	Flagship large language model designed for advanced reasoning, multilingual conversations, enterprise AI assistants, and complex workflow automation.
Sarvam-30B	Rs 2.5 / Rs 1.5 / Rs 10	Per 1M input / cached input/output tokens	Cost-efficient language model optimized for conversational AI, content generation, and enterprise productivity applications.
Sarvam Vision	Rs 0.5	Per page	Vision API for image understanding, document processing, OCR, and digitization workflows.
Text-to-Speech (Bulbul v3)	Rs 30	Per 10K characters	An advanced speech synthesis model that generates natural-sounding speech across multiple Indian languages and dialects.
Text-to-Speech (Bulbul v2)	Rs 15	Per 10K characters	Cost-effective text-to-speech model for voice assistants, IVR systems, and content narration.
Speech-to-Text	Rs 30	Per hour	Converts spoken audio into text with support for multiple Indian languages and accents.
Speech-to-Text with Diarization	Rs 45	Per hour	Transcribes audio while identifying and separating individual speakers in conversations.
Speech-to-Text & Translate	Rs 30	Per hour	Simultaneously transcribes and translates speech into target languages for multilingual communication.
Speech-to-Text, Translate & Diarization	Rs 45	Per hour	An end-to-end speech intelligence service that transcribes, translates, and identifies speakers in audio content.
Sarvam Translate V1	Rs 20	Per 10K characters	Neural machine translation service supporting high-quality translation across Indian languages.
Translate Mayura V1	Rs 20	Per 10K characters	Advanced multilingual translation model optimized for enterprise and government language workflows.
Transliterate	Rs 20	Per 10K characters	Converts text between different Indic scripts while preserving pronunciation and meaning.
Language Identification	Rs 3.5	Per 10K characters	Automatically detects and classifies the language of text inputs across multiple Indian languages.

Products Offering

Product	Description
Sarvam 30B and Sarvam 105B	Current multilingual language models for Indian-language, reasoning, agentic, and enterprise applications. Sarvam-M is deprecated.
Sarvam Agents	AI agents designed for customer support, workflow automation, and enterprise productivity use cases.
Sarvam Speech Platform	Speech-to-text and text-to-speech platform supporting numerous Indian languages and dialects.
Sarvam Translate	AI-powered translation platform enabling seamless communication across India's multilingual ecosystem.

On June 15, 2026, Sarvam announced a $234 million first close of a targeted $300 million Series B round at a $1.5 billion post-money valuation. The funding is intended to support frontier-model research, large-scale computing access, and enterprise and government deployment.

Sarvam also moved Saaras v3 into its speech-to-text stack and promoted Bulbul v3 to stable production status. The active chat models are Sarvam 30B and Sarvam 105B, while Sarvam-M is deprecated.

In February 2026, Sarvam AI announced landmark sovereign AI partnerships with the governments of Odisha and Tamil Nadu to accelerate the development of India’s AI infrastructure and public-sector AI capabilities. The collaboration focuses on building large-scale compute capacity, sovereign AI models tailored to Indian languages and regional contexts, and AI systems embedded directly into government services.

11. Kimi AI (Moonshot AI)

Parameter	Details
Company	Moonshot AI
AI Assistant	Kimi
Headquarters	Beijing, China
Core Focus	Multimodal AI, reasoning, coding, AI agents, long-context processing, and agentic workflows
Key Products	Kimi K3, Kimi K2.6, Kimi K2.5, Kimi Agent, Kimi Code, Kimi Docs, Kimi Sheets, Kimi API
Latest Flagship Model	Kimi K3
Website	https://www.kimi.com/

Kimi is an AI assistant developed by Moonshot AI, offering web search, deep thinking, multimodal reasoning, long-context conversations, agentic workflows, and developer APIs. Its recent product expansion has focused strongly on coding, office productivity, autonomous agents, and multi-agent execution.

Subscription Plans Pricing

Plan	Monthly Price	Annual Price (per month)	Agent Credits	Agent Concurrent Tasks	Agent Swarm (Beta)	Kimi Code Credits
Adagio	$0	—	6	1	—	—
Moderato	$19/month	$15/month	60	2	25 uses	1×
Allegretto	$39/month	$31/month	150	2	50 uses	5×
Allegro	$99/month	$79/month	360	4	120 uses	15×
Vivace	$199/month	$159/month	720	4	240 uses	30×

Products Offering

Product	Description
Kimi AI	AI assistant providing conversational AI, deep research, long-context processing, reasoning, document analysis, and productivity capabilities.
Kimi K3	Flagship 2.8-trillion-parameter model with native multimodal capabilities and a 1-million-token context window, designed for long-horizon coding, knowledge work, and reasoning.
Kimi K2.7 Code	Coding-focused AI model designed for software development, coding agents, long-context reasoning, and agentic programming workflows.
Kimi Code	Developer-focused coding service supporting terminal and IDE workflows, including CLI-based development and programming assistance.
Kimi Agent	Agentic AI capabilities for complex workflows, including research, document processing, presentations, spreadsheets, website generation, and other knowledge-work tasks.
Agent Swarm	Multi-agent capability that enables Kimi to coordinate multiple sub-agents for complex, parallel tasks.
Kimi Claw	AI agent capability available with selected paid membership tiers, extending Kimi's agentic functionality across supported platforms.
Kimi API	Developer platform providing API access to Kimi models through usage-based token pricing for integrating AI capabilities into applications and workflows.

In July 2026, Moonshot AI introduced Kimi K3, a 2.8-trillion-parameter, natively multimodal model with a 1-million-token context window, designed for long-horizon coding, knowledge work, and deep reasoning.

In April 2026, Moonshot AI released and open-sourced Kimi K2.6, delivering major upgrades in agentic coding, long-context reasoning, long-horizon execution, and Agent Swarm capabilities.

In January 2026, Moonshot AI launched Kimi K2.5, an open-source native multimodal model with improved coding, visual understanding, agent capabilities, and Agent Swarm, supporting up to 100 parallel sub-agents and 1,500 tool calls.

In January 2026, Kimi K2.5 expanded Kimi Agent's Office productivity capabilities, enabling end-to-end creation and processing of Word documents, PDFs, Excel spreadsheets, and presentations.

About the Author

Mamta Yadav is a Research Analyst with expertise in market intelligence, competitive analysis, and strategic research across global industries. She specializes in tracking market trends, evaluating competitive landscapes, and identifying emerging opportunities to deliver data-driven insights that support informed business decisions and long-term growth strategies.