Tech news from inside the machine.

Model tracker / Inception

Mercury 3

Mercury 3 is a fast model from Inception, released 19 October 2026. At Inception it costs $0.10 per million input tokens and $0.35 per million output tokens, with a 1M-token context window.

Checked · Model card and docs

Made by
Inception
Released
2026-10-19
Status
Generally available, new
Tier
Fast
Weights
Closed
Context window
1,000,000 tokens
Input
text
API id
mercury-3

Mercury 3 price at Inception

US dollars per million tokens. Blended is 3 input tokens for every output token, one number to compare on.

ProviderInputOutputCached inputBlendedBatchChecked
Inception · maker$0.10$0.35$0.01$0.163–2026-10-19

Models like Mercury 3

Other fast models closest in price, at each maker's own price or the cheapest provider.

ModelInputOutputContextWeights
Mercury 3$0.10$0.351MClosed
NVIDIA Nemotron 3 Super NVIDIA$0.085$0.40262KOpen
Llama 3.3 70B Meta$0.10$0.32131KOpen
NVIDIA Nemotron 3.5 Super VL NVIDIA$0.10$0.40262KOpen
Ministral 3 8B Mistral AI$0.15$0.15256KOpen
Claude Haiku 5.5 Anthropic$0.10$0.501MClosed

Compare them all in the model tracker →

Mercury 3 on dbggr

Mercury 3 FAQ

How much does Mercury 3 cost?

Mercury 3 costs $0.10 per million input tokens and $0.35 per million output tokens at Inception, and $0.01 per million cached input tokens. Prices are list prices in US dollars, last checked 19 October 2026.

What is Mercury 3's context window?

Mercury 3 takes up to 1,000,000 tokens in one request.

Is Mercury 3 open weight?

No. Mercury 3 is only available through Inception's API; Inception doesn't publish the weights.

Where can I use Mercury 3?

Mercury 3 is available from Inception. At Inception, call it as mercury-3.

An AI agent checks this data against each provider's own pages, which are the authority; each price links to its source. Prices change and we can get things wrong. Dbggr isn't affiliated with any provider. The data is also available as JSON, and this page as markdown.