Together AI: New Provisioned Throughput pricing + GPU rate cuts
What happened
Between the Jul 6, 2026 and Jul 13, 2026 captures of Together AI's pricing page, Pulse recorded one change: Added Provisioned Throughput (PTU) reserved-capacity pricing ($0.05/PTU/min, MiniMax M3 & GLM-5.2); Dedicated Inference on-demand GPU rates decreased (H100 $6.49->$5.49/hr, HGX B200 $11.95->$8.99/hr). In detail: Added Provisioned Throughput (PTU) pricing ($0.05/PTU/min, MiniMax M3 & GLM-5.2); Dedicated Inference on-demand GPU rates cut (H100 $6.49->$5.49/hr, HGX B200 $11.95->$8.99/hr). Pulse files it as a packaging change (package change) with a neutral direction. Pulse's latest capture (Oct 26, 2026) lists Post-Training at $9 per GPU-hour (B200, promo), Serverless Inference at $0.14 per 1M input tokens, Dedicated Inference at $5.49 per GPU-hour (H100, on-demand) + $8.99 per GPU-hour (B200, on-demand), GPU Clusters (On-demand) at $3.99 per hour (H100) and Fine-Tuning at $0.34 per 1M tokens (LoRA SFT, cheapest models) + $0.38 per 1M tokens (Full SFT, cheapest models).
It lands 13 days after Together AI's previous change, "Together AI drops GLM-5 from serverless inference pricing" (Jun 30, 2026). Before that Pulse recorded "Together AI: Inkling model price drop ($1.20→$1.00)" (Jul 18, 2026), "Together AI: Added Inkling & LFM2.5-8B-A1B models to Serverless Inference" (Jul 17, 2026) and "Together AI: new flagship models added, Llama 4 Maverick + Kimi K2.5 removed" (Jul 4, 2026). Together AI has been tracked since 2025-12-23.
In the same window Google, Anthropic, Huggingface, Mistral AI and OpenAI also changed their AI & Machine Learning pricing pages. Those moves are listed further down, and the Pulse Agent can compare them side by side.
- Provisioned Throughput, Dedicated Inference · Packaging−15%$6.49$5.49/hr
| Plan | Field | Before | After | Change |
|---|---|---|---|---|
| Provisioned Throughput, Dedicated Inference | Packaging | $6.49 | $5.49/hr | −15% |
Values as written on Together AI's public pricing page in the two captures Pulse compared; where the page states no number, the cells give the wording before and after.
Evidence
The two captures Pulse comparedFull-page captures of Together AI's pricing page (www.together.ai) from Jul 6, 2026 and Jul 13, 2026; the change is marked in the lower part of the page, where provisioned throughput, dedicated inference packaging reads $6.49 and then $5.49/hr.
Sign in to see the evidence
The two full-page captures with this change marked up are free with an account.
Together AI, the last 90 days
Full history- Jul 20, 2026improvedTogether AI: 3 new models added to Serverless Inference pricing
- Jul 18, 2026decreaseTogether AI: Inkling model price drop ($1.20→$1.00)
- Jul 17, 2026improvedTogether AI: Added Inkling & LFM2.5-8B-A1B models to Serverless Inference
- Jul 13, 2026packagingTogether AI: New Provisioned Throughput pricing + GPU rate cuts · this change
- Jun 30, 2026packagingTogether AI drops GLM-5 from serverless inference pricing
- Jul 4, 2026packagingTogether AI: new flagship models added, Llama 4 Maverick + Kimi K2.5 removed
- Jun 22, 2026decreaseTogether AI slashes GPU and AI inference prices up to 25%
- Jun 19, 2026packagingTogether AI reveals cached input pricing for GLM-5.1 and Qwen3.5
- Jun 10, 2026increaseTogether AI raises serverless inference prices across Llama, Qwen models
- Jun 2, 2026decreaseTogether AI cuts Qwen3 inference prices 50%, adds caching
Across AI & Machine Learning in the same window
All AI & Machine Learning changes- GoogleJul 6Google Vertexai delays Agent Gateway billing to July 13, extends free window
- AnthropicJul 6Anthropic launches Sonnet 5, cuts price to $2/$10 per MTok through Aug 31
- HuggingfaceJul 3Enterprise SCIM provisioning & Team SSO OIDC added + 2 more changes
- Mistral AIJun 2Mistral AI adds Task scheduling, Skills, and Workflows in major overhaul
- OpenAIMay 11OpenAI API raises GPT-Realtime-2 text output 50%, adds translate
Questions about this change
Did Together AI raise prices in July 2026?
No list-price increase was recorded between Jul 6, 2026 and Jul 13, 2026. The change Pulse detected was a packaging change: package change.
What changed on Together AI's pricing page between Jul 6, 2026 and Jul 13, 2026?
Provisioned Throughput, Dedicated Inference packaging: $6.49 → $5.49/hr (−15%). Pulse compared the captures it took on Jul 6, 2026 and Jul 13, 2026; the before/after table above lists each item as written on the page.
How often does Together AI change its pricing?
Pulse recorded 8 other pricing-page changes for Together AI in the 90 days before this one; the previous change was "Together AI drops GLM-5 from serverless inference pricing" on Jun 30, 2026; Together AI has been tracked since 2025-12-23.