One agent session
1M input tokens and 100K output tokens, no caching.
- Input1M x $0.20 = $0.20
- Output0.1M x $0.75 = $0.075
$0.275per session
Inception - Closed weights
Mercury 2.5 pricing and specs
Mercury 2.5 is a closed-weights model from Inception that costs $0.20 per 1M input tokens and $0.75 per 1M output tokens on the first-party API, with a 260K-token context window.
Last verified from docs.inceptionlabs.ai. All sources
- Input / 1M
- $0.20
- Output / 1M
- $0.75
- Context
- 260K tokens
- Weights
- Closed
Pricing per 1M tokens
| Item | Price / 1M tokens |
|---|---|
| Input | $0.20 |
| Output | $0.75 |
| Cache read | $0.02 |
Note: Pricing records the standard list price. A launch promotion is live on 2026-10-17: 80% off, billed at $0.04 input, $0.004 cached input and $0.15 output per 1M tokens (the docs and models page show the list price struck through; the OpenAPI model object reports the promo price; no end date is stated). Diffusion LLM (dLLM), text in/out, reasoning with low/medium/high reasoning_effort, tool calling, structured outputs. Context window stated as 260K on the docs table and 260000 in the API model object. No separate cache-write or batch price is published. Gateway id confirmed on the Vercel AI Gateway model page; gateway prices not checked against first-party.
Specs
- Context window
- 260,000 tokens
- Max output
- 65,536 tokens
- Release date
- Provider
- Inception
- API model id
mercury-2.5- Vercel AI Gateway id
inception/mercury-2.5- Weights
- Closed weights
What Mercury 2.5 costs in practice
Worked from the standard list rates above, with the math shown so you can swap in your own token counts.
A chat app
10,000 requests a day at 2K input and 500 output tokens each.
- Input20M x $0.20 = $4.00
- Output5M x $0.75 = $3.75
$7.75per day
$232.50 per 30-day month
The same session at a 90% cache hit
90% of the input is read from the prompt cache. The one-time cache write premium is not included.
- Cached input0.9M x $0.02 = $0.018
- Uncached input0.1M x $0.20 = $0.02
- Output0.1M x $0.75 = $0.075
$0.113per session
Compare with
| Model | Input / 1M | Output / 1M |
|---|---|---|
| Mercury 2.5 (this page) | $0.20 | $0.75 |
| Mercury 2Inception | $0.25 | $0.75 |
| Mercury Edit 2Inception | $0.25 | $0.75 |
| MiMo-V2.6-ProXiaomi MiMo | $0.435 | $0.87 |
| Codestral 25.08Mistral AI | $0.30 | $0.90 |
Related reading
- Mercury 2 Developer Guide: Building With a Diffusion LLM in ProductionMercury 2 developer guide: the OpenAI-compatible API, tool use, structured outputs, reasoning levels, and when a diffusion LLM beats an autoregressive one.
- Mercury 2: The LLM That Doesn't Generate Like an LLMMercury 2 from Inception Labs is a diffusion reasoning LLM that refines outputs in parallel at over 1,000 tokens per second. How it works and where it fits.
- Diffusion Language Models: How Mercury Changed the LLM Speed GameDiffusion language models like Inception Labs' Mercury draft a whole response and refine it instead of going token by token. See why that makes them so fast.
FAQ
How much does Mercury 2.5 cost per million tokens?
Mercury 2.5 costs $0.20 per 1M input tokens and $0.75 per 1M output tokens on Inception's first-party API, per docs.inceptionlabs.ai on October 17, 2026. Per 1M tokens, cache reads cost $0.02.
What is Mercury 2.5's context window?
Mercury 2.5 has a 260,000-token context window and returns up to 65,536 output tokens per response.
Are Mercury 2.5's weights open?
No. Mercury 2.5 is a closed-weights model. You call it through Inception's API as mercury-2.5 or through Vercel AI Gateway as inception/mercury-2.5.
When was Mercury 2.5 released?
Inception released Mercury 2.5 on September 8, 2026.
Sources
Last verified from docs.inceptionlabs.ai.
Newsletter
Get model price changes as they ship
One email per week. New model releases, price cuts and what they mean for your bill.