fal raises H100 GPU rate to $2.49/hr, adds GB200, doubles MiniMax video price
What happened
Between the Oct 1, 2026 and Oct 2, 2026 captures of fal's pricing page, Pulse recorded one change: fal Serverless GPU rates rose (H100 $1.89->$2.49/hr, B300 $4.49->$5.99); GB200 GPU added; MiniMax H3 Max video per-second prices doubled ($0.025->$0.05). Pulse files it as a pricing change (price increase) with an upward direction. Pulse's latest capture (Oct 20, 2026) lists Audio Model APIs at $0.1 per 1,000 characters (MiniMax Speech 2.8 HD) + $0.2 per 100 seconds (ElevenLabs Sound Effects v2), 3D Model APIs at $0.08 per generation (Hunyuan Motion 1B) + $0.01 per credit (Tripo H3.1), Serverless Compute at $6 per H200 GPU hour + $4 per RTX PRO 6000 GPU hour, Video Model APIs at $0.05 per second of video output (MiniMax H3 Max) + $0.07 per second of video output (Kling 2.5 Turbo Pro) and Image Model APIs at $1 per unit (GPT Image 2) + $0.5 per 100 megapixels (Z Image Turbo).
It lands one day after fal's previous change, "fal publishes Audio and 3D model API pricing, expands video/image tables to 10 endpoints" (Oct 1, 2026). Before that Pulse recorded "fal: H100 GPU list price increase ($3.99→$4.50/h)" (Aug 5, 2026). fal has been tracked since 2026-07-15.
In the same window Together AI, Anthropic, Modal and Runway 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 GPU · Price+32%$1.89$2.49/hr
| Plan | Field | Before | After | Change |
|---|---|---|---|---|
| Serverless GPU | Price | $1.89 | $2.49/hr | +32% |
Values as written on fal'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 fal's pricing page (fal.ai) from Oct 1, 2026 and Oct 2, 2026; the change is marked near the top of the page, where serverless gpu price reads $1.89 and then $2.49/hr.
Sign in to see the evidence
The two full-page captures with this change marked up are free with an account.
fal, the last 90 days
Full history- Oct 12, 2026improvedfal adds Flux 3 Image editing at $0.024 per megapixel to Model API
- Oct 9, 2026improvedfal adds Flux 3 image editing to rate card at $0.024 per megapixel
- Oct 5, 2026improvedfal hikes GPU rates up to 33%, adds GB200 tier and Nano Banana 2/Pro models
- Oct 2, 2026increasefal raises H100 GPU rate to $2.49/hr, adds GB200, doubles MiniMax video price · this change
- Oct 1, 2026packagingfal publishes Audio and 3D model API pricing, expands video/image tables to 10 endpoints
- Aug 5, 2026increasefal: H100 GPU list price increase ($3.99→$4.50/h)
Across AI & Machine Learning in the same window
All AI & Machine Learning changes- Together AIOct 2Together AI adds Tev1 4B Experimental at $0.04 input with free output tokens
- AnthropicSep 28Anthropic API: Opus 5.5 launches, replaces Opus 5 as flagship
- ModalSep 14Modal renames Enterprise SSO from Okta to standards-based SAML
- RunwaySep 7Runway slashes Pro storage to 100GB, adds Agentic collaborator and Brand Kits
Questions about this change
Did fal raise prices in October 2026?
Yes. Pulse captured fal's pricing page on Oct 1, 2026 and again on Oct 2, 2026 and recorded 1 price increase: Serverless GPU: $1.89 → $2.49/hr (+32%).
What changed on fal's pricing page between Oct 1, 2026 and Oct 2, 2026?
Serverless GPU price: $1.89 → $2.49/hr (+32%). Pulse compared the captures it took on Oct 1, 2026 and Oct 2, 2026; the before/after table above lists each item as written on the page.
How often does fal change its pricing?
Pulse recorded 2 other pricing-page changes for fal in the 90 days before this one; the previous change was "fal publishes Audio and 3D model API pricing, expands video/image tables to 10 endpoints" on Oct 1, 2026; fal has been tracked since 2026-07-15.