Together AI adds Qwen3.8 Flash at $0.15/1M tokens, drops Gemma 3n E4B
What happened
Between the Aug 31, 2026 and Sep 1, 2026 captures of Together AI's pricing page, Pulse recorded one change: Serverless Inference model catalog updated: new model 'Qwen3.8 Flash' added ($0.15/$0.47 per 1M tokens); 'Gemma 3n E4B Instruct' model removed (was $0.06/$0.12 per 1M tokens). Pulse files it as a product change (feature 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 3 days after Together AI's previous change, "Together AI cuts Qwen3.8-2.4T-A95B inference prices and adds GLM-5.3" (Aug 29, 2026). Before that Pulse recorded "Together AI adds GLM-5.3-Flash, drops DeepSeek V4 Pro and Kimi K2.7 Code" (Aug 28, 2026), "Together AI unveils Qwen3.8-2.4T-A95B, Muse Glimmer, DeepSeek V4 Pro pricing on " (Aug 17, 2026) and "Together AI: HGX H100 GPU price increase + Kimi K3 & Inkling Small models added" (Aug 4, 2026). Together AI has been tracked since 2025-12-23.
In the same window OpenAI, Mistral AI, Lambda Labs, Anthropic and Google also changed their AI & Machine Learning pricing pages. Those moves are listed further down, and the Pulse Agent can compare them side by side.
- Serverless Inference · FeatureFeature change
Serverless Inference model catalog updated: new model 'Qwen3.8 Flash' added ($0.15/$0.47 per 1M tokens); 'Gemma 3n E4B Instruct' model removed (was $0.06/$0.12 per 1M tokens).
| Plan | Field | Before | After | Change |
|---|---|---|---|---|
| Serverless Inference | Feature | Serverless Inference model catalog updated: new model 'Qwen3.8 Flash' added ($0.15/$0.47 per 1M tokens); 'Gemma 3n E4B Instruct' model removed (was $0.06/$0.12 per 1M tokens). | Feature change | |
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 Aug 31, 2026 and Sep 1, 2026; the change is marked near the top of the page.
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- Sep 14, 2026increaseTogether AI: Qwen3.7-Max price hike + new DeepSeek V4.1 Flash model
- Sep 12, 2026packagingTogether AI replaces 3 fine-tuning size tiers with 29 per-model price rows
- Sep 11, 2026increaseTogether AI hikes Qwen3.7-Max pricing 60% and cuts LFM2.5-8B model
- Sep 1, 2026improvedTogether AI adds Qwen3.8 Flash at $0.15/1M tokens, drops Gemma 3n E4B · this change
- Aug 29, 2026improvedTogether AI cuts Qwen3.8-2.4T-A95B inference prices and adds GLM-5.3
- Aug 28, 2026improvedTogether AI adds GLM-5.3-Flash, drops DeepSeek V4 Pro and Kimi K2.7 Code
- Aug 17, 2026improvedTogether AI unveils Qwen3.8-2.4T-A95B, Muse Glimmer, DeepSeek V4 Pro pricing on ...
- Aug 4, 2026improvedTogether AI: HGX H100 GPU price increase + Kimi K3 & Inkling Small models added
- Jul 28, 2026improvedTogether AI: Kimi K3 model added to pricing
- Jul 27, 2026packagingTogether AI: LFM2.5-8B-A1B model removed from Serverless Inference
Across AI & Machine Learning in the same window
All AI & Machine Learning changes- OpenAIAug 31OpenAI API Business: new $100/mo Premium seat tier + 200-employee cap
- Mistral AIAug 17Mistral AI folds Education into Pro, adds bundled API credits to entry plans
- Lambda LabsJul 27Lambda Labs gates HGX B200 clusters behind sales, keeps $9.86–$8.87/hr rates
- AnthropicJul 23Anthropic: Fable model + context window disclosed in comparison table
- GoogleJul 6Google Vertexai delays Agent Gateway billing to July 13, extends free window
Questions about this change
Did Together AI raise prices in September 2026?
No list-price increase was recorded between Aug 31, 2026 and Sep 1, 2026. The change Pulse detected was a product change: feature change.
What changed on Together AI's pricing page between Aug 31, 2026 and Sep 1, 2026?
Serverless Inference model catalog updated: new model 'Qwen3.8 Flash' added ($0.15/$0.47 per 1M tokens); 'Gemma 3n E4B Instruct' model removed (was $0.06/$0.12 per 1M tokens). Pulse compared the captures it took on Aug 31, 2026 and Sep 1, 2026; the before/after table above lists each item as written on the page.
How often does Together AI change its pricing?
Pulse recorded 16 other pricing-page changes for Together AI in the 90 days before this one; the previous change was "Together AI cuts Qwen3.8-2.4T-A95B inference prices and adds GLM-5.3" on Aug 29, 2026; Together AI has been tracked since 2025-12-23.