Mercury API Pricing
Inception builds diffusion LLMs — models that generate tokens in parallel rather than one after another. The Mercury family is the only architecture of its kind priced in this catalogue, and the company sells it on throughput: over 1,000 tokens per second, or as its own marketing puts it, frontier quality at five times the speed.
Mercury 3 LAUNCHED OCT 2026
Mercury 3 is Inception's flagship diffusion LLM, released 2026-10-19 on its own API and through OpenRouter, Baseten and Vercel.
Mercury Edit 2 LAUNCHED AUG 2026
Code-editing sibling of Mercury 2 at the identical rate card ($0.25 / $0.025 / $0.75).
Mercury Voice LAUNCHED SEP 2026
⚠ Enterprise customers only - Inception enables it per organisation on request through [email protected]. Mercury Voice is Inception's diffusion LLM tuned for voice agents - the language model…
Mercury 2.5 LAUNCHED SEP 2026
Mercury 2.5 is Inception's diffusion LLM for fast reasoning, chat and agents, generally available on its own API since 2026-09-08.
Mercury 2 LAUNCHED AUG 2026
Inception's previous-generation reasoning dLLM, kept for existing customers since Mercury 2.5 shipped.
Mercury 2.5 Preview LAUNCHED AUG 2026
Inception prices Mercury 2.5 Preview at $0.20/M input, $0.02/M cached input and $0.75/M output on its own models page, and describes it as "our most intelligent reasoning dLLM, at our lowest price"…
| Model | Input /M | Output /M | Cached | Context | Max output | Vision | Tools | Tier |
|---|---|---|---|---|---|---|---|---|
| Mercury 3 FLAGSHIP | $0.20 | $0.70 | $0.02−90% | 1M | — | ✗ | ✓ | Frontier |
| Mercury Edit 2 | $0.25 | $0.75 | $0.025−90% | 32K | — | ✗ | ✓ | Active |
| Mercury Voice RESTRICTED | $0.40 | $1.50 | $0.04−90% | 128K | Enterprise customers only - Inception enables it per organisation on request through [email protected] | ✗ | ✗ | Restricted |
| Mercury 2.5 DEPRECATED | $0.20 | $0.75 | $0.02−90% | 260K | — | ✗ | ✓ | Deprecated |
| Mercury 2 DEPRECATED | $0.25 | $0.75 | $0.025−90% | 128K | — | ✗ | ✓ | Deprecated |
| Mercury 2.5 Preview PREVIEW (ARCHIVED) | — | — | — | Withdrawn | → Mercury 3 | ✗ | ✗ | Withdrawn |
Pricing across the lineup.
The Mercury story so far.
Prices are read from Inception's own developer table and models page. Company history is from the launch coverage and the lead investor's announcement, cited as reported rather than vendor-stated.
platform.inceptionlabs.ai — API keys and playground. The endpoint is OpenAI-compatible, so adopting it is a base-URL and model-name change rather than a rewrite. Every new account is granted 100 million free tokens.
DOCSThe developer table that prices Mercury 3, Mercury Voice, Mercury Decide, Mercury 2.5, Mercury 2 and Mercury Edit 2, with endpoints, context windows and max output per model, and launch discounts shown against list. This is the source for the rates on this page.
RESELLERResells Mercury 3 alongside Inception's own API, Baseten and Vercel; it was the only route to the Mercury 2.5 preview. A reseller sets its own margin, so the rate you pay there need not equal the vendor's published figure.
Inception came out of stealth in February 2025 and is based in Palo Alto. It was founded by three academics — Stefano Ermon of Stanford, who co-invented diffusion models, Aditya Grover of UCLA and Volodymyr Kuleshov of Cornell. In November 2025 it raised a $50 million seed led by Menlo Ventures, with Mayfield, Innovation Endeavors, Microsoft's M12, Snowflake Ventures, Databricks and Nvidia's NVentures participating. Those company facts come from launch coverage and the investors' own announcements, not from a vendor pricing page.
The technical bet is what makes this provider worth a page. Every other model in this catalogue is autoregressive: it produces one token, then the next, conditioned on what came before. A diffusion LLM starts from noise and refines a whole span in parallel. The practical consequence is throughput — Inception quotes over 1,000 tokens per second for Mercury 2 — which is why the company's own framing is speed rather than price.
And that framing matters when reading the rates. At $0.20–$0.25 per million input tokens Mercury is not a budget outlier: it lands in the same band as the mainstream cheap tiers from the large labs, above the cheapest Chinese rows. Buying Mercury to cut a token bill is the wrong reason. Buying it because a task is latency-bound — voice, autocomplete, an agent loop where a slow frontier model is unusable at any price — is the argument the vendor is actually making, and the one worth testing.
The line moves fast. Mercury 2.5 reached general availability on September 8, 2026 at 80% off, and six weeks later Mercury 3 replaced it: a 1M-token window, a $0.20/$0.70 list card and 50% off at launch, again with no end date. Mercury 2 and 2.5 closed to new users the same day, with access for existing users until November 15, 2026, and Mercury 2.5's discount ended without notice. The site prices every Mercury model at list and shows a discount beside it, because a discount with no end date can stop on any day.
Google (Gemini Flash)
The mainstream fast tier most teams reach for first, with vision and a far larger ecosystem. Mercury's pitch against it is raw tokens per second, not capability breadth or price.
PRICE PEERAnthropic (Haiku)
The model Inception benchmarks Mercury 3 against, from a frontier lab, with vision, caching and a much wider tool ecosystem, and a lower list card than Mercury 3. If latency is not the binding constraint, this is the safer default.
CHEAPERAlibaba (Qwen Flash)
Materially cheaper per token and multimodal, but China-hosted, which is a procurement question Inception does not raise. Mercury does not win on price against this tier.
CODE PEERKwaipilot (KAT Coder)
Coding-specialised rows in a comparable band, aimed at the same IDE and agent workloads Mercury Edit 2 targets.
OPEN-WEIGHT GIANTAlibaba (Qwen)
The Qwen (Tongyi Qianwen) model family from Alibaba Cloud's Tongyi Lab. First released in 2023, Qwen is the most-downloaded open-weight model family in the world — most tiers ship under Apache 2.0 on Hugging Face and ModelScope, while the proprietary Max tier is served through Alibaba Cloud's Model Studio.
FRONTIER LABOpenAI
The lab behind ChatGPT and the GPT family. Founded as a nonprofit in 2015 by Sam Altman, Elon Musk, Ilya Sutskever, Greg Brockman and others; restructured into a capped-profit LLC in 2019 and finally into a public-benefit corporation on October 28, 2025. Ships the largest commercial model lineup, the largest consumer AI product, and — after the April 2026 funding round — the highest-valued private AI company in history at ~$852B.
Frequently asked.
Practical questions about Mercury pricing, the free allowance, and what a diffusion LLM does and does not change about your bill.
Q · 01 Which Mercury model should I start with? +
$0.20/M input and $0.70/M output with a 1M window, and is 50% off at launch. Mercury 2 and Mercury 2.5 are closed to new users and keep existing users only until November 15, 2026. Mercury Voice is for enterprise voice agents and is enabled through sales. Mercury Edit 2 is for autocomplete and next-edit tooling, not chat.Q · 02 Is Mercury actually cheap? +
$0.20 input and $0.70 output sit above GPT-6 Luna and Claude Haiku 5.5, the two models Inception benchmarks it against, and above the cheapest Chinese rows. While the 50% launch discount runs, $0.10/$0.35 ties those two on input and undercuts them on output, but Inception gives no end date, and the 80% discount on Mercury 2.5 ended without one. What the company sells is speed - over 1,000 tokens per second - so the honest comparison is cost per completed task under a latency budget.