Together AI Annualized Revenue Run-Rate
FY2024-Feb 2026
Annualized Run-Rate (USD M)
Updated at 2026-06
Together AI Funding Rounds and Valuation Trajectory
| Round (date, lead) | Amount Raised | Post-Money Valuation | Notable Investors |
|---|---|---|---|
| Seed (2023-05, Lux Capital) | $20M | n/a | Lux Capital, SCB10X, Definition Capital, Cadenza, Susa Ventures |
| Series A (2023-11, Kleiner Perkins) | $102.5M | ~$1B (Tracxn) | Kleiner Perkins, NVIDIA, Emergence Capital, NEA, Prosperity7, Greycroft, Lux Capital |
| Series A extension (2024-03, Salesforce Ventures) | $106M | $1.25B | Salesforce Ventures, Coatue, Kleiner Perkins, Lux Capital, NEA |
| Series B (2025-02, General Catalyst + Prosperity7) | $305M | $3.3B | General Catalyst, Prosperity7 (Saudi Aramco), Salesforce Ventures, NVIDIA, Kleiner Perkins, Coatue |
| New round (2026, RUMORED: reported closed April 2026 per The Information, no primary Together announcement) | ~$1B (reported) | $7.5B (reported) | Prosperity7 cited as lead per reports; round details unconfirmed by Together |
Updated at 2026-06
Together AI GPU Cluster Pricing by NVIDIA Generation
| NVIDIA GPU (architecture) | On-demand HGX list ($/GPU-hr) | Self-service Instant Cluster band ($/GPU-hr) | Fabric / status |
|---|---|---|---|
| H100 (Hopper) | $5.49 | $1.76 to $2.39 | NVLink + InfiniBand; GA |
| H200 (Hopper) | $6.79 | $3.15 to $3.79 | NVLink + InfiniBand; GA |
| B200 (Blackwell) | $9.95 | $4.00 to $5.50 | NVLink + InfiniBand; early-access GA |
| GB200 (Grace Blackwell) | Contact us | Not listed | Quoted on request; newest tier |
Updated at 2026-06
Together AI Open-Model Cloud: Scale, Catalog, and Flagship Customers
| Metric / customer group | Figure or detail | Significance for Together |
|---|---|---|
| Developers and AI-native companies | 450,000+ (company-stated) | Breadth of the open-model developer funnel that feeds per-token inference |
| Open models served (serverless inference) | 100+ open-weight models | Llama, DeepSeek, Qwen, gpt-oss; the open-source catalog is the core differentiator |
| Power capacity | ~200 MW across North America | Underlying GPU fleet scale; shifting toward owned vs subleased capacity |
| Blended gross margin | ~45% | Rentals carry capital intensity; owning GPUs is the margin lever |
| Coding / agents customers | Cursor, Cognition, Decagon, Vercept | Latency-sensitive agentic workloads on Together inference |
| Voice and media customers | Cartesia, ElevenLabs, Pika, Leonardo, KREA, Hedra | Real-time generative media built on Together GPUs |
| Enterprise and platform customers | Salesforce, Zoom, Quora, Zoho, Arcee, Nous | Enterprise inference plus model-builder customers |
| Together-authored open releases | RedPajama dataset (2023), DeepCoder-14B (2025) | Ecosystem anchors that drive open-model mindshare back to the cloud |
Updated at 2026-06
Together AI Revenue Mix: GPU Cluster Rentals vs Inference API
End 2025 (~$618M, est. split)-Feb 2026 (~$1B, est. split)
Updated at 2026-06
Together AI in AI Software
OpenAI
$PRIVATE
Largest AI lab by revenue. Approximately $13B annualized run-rate revenue (mid-2025) from a mix of ChatGPT consumer subscriptions, ChatGPT Enterprise / Team, the OpenAI API, and Microsoft Azure OpenAI revenue share. ChatGPT reaches roughly 800M weekly active users (late 2025). Builder of the GPT family (GPT-4o, GPT-4.5, GPT-5, o-series reasoning models) and Sora video. Microsoft is the dominant compute and distribution partner; SoftBank led the $40B March 2025 primary round at $300B.
