Together AI slashes GPU and AI inference prices up to 25%
What happened
Between the Jun 15, 2026 and Jun 22, 2026 captures of Together AI's pricing page, Pulse recorded 2 changes: GPU Clusters on-demand rates cut across all hardware: H100 $5.49→$4.79, H200 $6.79→$5.99, B200 $9.95→$8.19/hour. Reserved rates also reduced 12–25% across H100, H200, and B200 tiers; DeepSeek V4 Pro serverless inference prices reduced: input from $2.10 to $1.74/1M tokens (~17% cut) and output from $4.40 to $3.48/1M tokens (~21% cut). Cached input unchanged at $0.20. Pulse files it as a pricing change (price decrease) with a downward direction. Pulse's latest capture (Oct 21, 2026) lists Managed Storage at $0.16 per GiB/month, Provisioned Throughput at $0.05 per PTU/minute, Sandbox (Code Interpreter) at $0.03 per session (60 min), GPU Clusters (On-demand) at $3.99 per hour (H100) and GPU Clusters (Reserved) at $3.69 per hour (H100, 3-30 days).
It lands 3 days after Together AI's previous change, "Together AI reveals cached input pricing for GLM-5.1 and Qwen3.5" (Jun 19, 2026). Before that Pulse recorded "Together AI raises serverless inference prices across Llama, Qwen models" (Jun 10, 2026), "Together AI cuts Qwen3 inference prices 50%, adds caching" (Jun 2, 2026) and "Together AI reveals H200 GPU rates and adds Qwen3.7-Max inference" (May 29, 2026). Together AI has been tracked since 2025-12-23.
In the same window Anthropic, Huggingface, Google, 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.
- B200 · Price−13%$5.49$4.79
- Serverless Inference · Price−17%$2.10$1.74
| Plan | Field | Before | After | Change |
|---|---|---|---|---|
| B200 | Price | $5.49 | $4.79 | −13% |
| Serverless Inference | Price | $2.10 | $1.74 | −17% |
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 Jun 15, 2026 and Jun 22, 2026; the change is marked around the middle of the page, where b200 price reads $5.49 and then $4.79.
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 13, 2026packagingTogether AI: New Provisioned Throughput pricing + GPU rate cuts
- 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% · this change
- Jun 19, 2026packagingTogether AI reveals cached input pricing for GLM-5.1 and Qwen3.5
- Jun 10, 2026improvedTogether AI raises serverless inference prices across Llama, Qwen models
- Jun 2, 2026decreaseTogether AI cuts Qwen3 inference prices 50%, adds caching
- May 29, 2026packagingTogether AI reveals H200 GPU rates and adds Qwen3.7-Max inference
- Apr 24, 2026increaseTogether AI Triples Gemma 3n E4B Pricing to $0.06/$0.12 per 1M Tokens
- Apr 13, 2026improvedTogether AI adds 11 models including Gemma 4 and raises GLM-5 prices up to 40%
Across AI & Machine Learning in the same window
All AI & Machine Learning changes- AnthropicJun 19Anthropic API shifts Opus 4.8 focus to enterprise agentic coding
- HuggingfaceJun 8Huggingface reframes PRO pitch from crowd count to value promise
- GoogleJun 8Google Cloud reveals AI Commerce Search pricing at $2.50/1K
- 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 June 2026?
No — prices went down. Between Jun 15, 2026 and Jun 22, 2026 Together AI's pricing page showed 2 price cuts: B200: $5.49 → $4.79 (−13%); Serverless Inference: $2.10 → $1.74 (−17%).
What changed on Together AI's pricing page between Jun 15, 2026 and Jun 22, 2026?
B200 price: $5.49 → $4.79 (−13%); Serverless Inference price: $2.10 → $1.74 (−17%). Pulse compared the captures it took on Jun 15, 2026 and Jun 22, 2026; the before/after table above lists each item as written on the page.
How often does Together AI change its pricing?
Pulse recorded 9 other pricing-page changes for Together AI in the 90 days before this one; the previous change was "Together AI reveals cached input pricing for GLM-5.1 and Qwen3.5" on Jun 19, 2026; Together AI has been tracked since 2025-12-23.