How much does Mercury 2 cost?

Mercury 2 costs $0.25 per 1 million input tokens and $0.75 per 1 million output tokens. This pricing is set by Inception Labs.

What tokenizer does Mercury 2 use?

Mercury 2 uses the cl100k_base tokenizer. This tokenizer has a vocabulary of ~100,000 tokens and is widely used across many models.

What is Mercury 2's context window?

Mercury 2 supports a context window of 128K tokens (128,000 tokens). The maximum output length is 50K tokens.

How does Mercury 2 compare to other models?

You can compare Mercury 2's pricing and capabilities against other LLMs in the comparison table above, or visit the full pricing page for a side-by-side view of all models.

How do I count tokens for Mercury 2?

Use the token counter at the top of this page or visit the homepage calculator. Paste your text and it will show the exact token count and cost for Mercury 2 using the cl100k_base tokenizer.

Mercury 2

Last verified September 7, 2026 · Inception Labs pricing ↗

$0.250/1M input · $0.750/1M output · 128K context · Inception Labs

Count Tokens

Tokens

—

Input Cost

—

Output Cost

—

Tokens

—

Input Cost

—

Output Cost

Estimate Monthly Cost

Monthly Cost Estimator

Quick:

API calls / dayAvg input tokens / callAvg output tokens / call

<$0.0001/mo

Pick a preset above or enter custom usage

Alternatives to Mercury 2

Cheaper Alternative

Qwen3 Coder Next

$0.120 in · $0.800 out · Alibaba

8% cheaper with same or larger context

Premium Upgrade

Nemotron 3 Super 120B

$0.300 in · $0.800 out · NVIDIA

Upgrade with 1000K context

Same Provider

Mercury 2.5 (Preview)

$0.200 in · $0.750 out · Inception Labs

5% cheaper in Inception Labs ecosystem

Pricing Details

Inception LabsFirst diffusion-based reasoning LLM. $0.25/$0.75 confirmed on three Inception surfaces that agree: the models page, the docs pricing table and its own /v1/models endpoint. Cached input $0.025/1M; cache writes are free ($0). Still sold, and as of September 7, 2026 it is the only model Inception's own API serves. Speed has three live figures and they are not interchangeable: 1,009 tok/s on NVIDIA Blackwell per Inception's launch post (vendor lab claim), 709.9 output tok/s measured by Artificial Analysis (2nd of 177), and 233 tok/s P50 across providers on OpenRouter's live traffic. We previously carried 788 tok/s attributed to Artificial Analysis; that page now reads 709.9 and we could not reconstruct the earlier capture, so this row is corrected to the current figure and dated. AA Intelligence Index 15. AIME 2025 91.1, GPQA Diamond 73.6, HumanEval 89.4 circulate widely but we could not find them on any Inception first-party page, so treat them as unverified. Verbosity caveat: produces ~2.6x output tokens of median model on AA suite. Sibling Mercury Edit 2 (FIM/NextEdit, 32K, 8,192 max output) shares the same rate card. Mercury Coder is absent from Inception's site, docs and API and from OpenRouter's catalogue. Estimated tokens (different tokenizer).

Input / 1M tokens

$0.25

Output / 1M tokens

$0.75

Context Window

128K

Max Output

50K

Full Pricing Table Token Calculator

Price History

No price change recorded since 2026-08-11, when this site started tracking this model; it launched on 2026-05-12.

Frequently Asked Questions

Common questions about Mercury 2 pricing and usage