Mercury 2: benchmark scores, price and ranking
Mercury 2 from Inception Labs ranks #614 of 743 LLMs on the LLMs Tiger aggregate benchmark score (44.9), based on 5 benchmark results captured up to 2026-10-13.
- Organisation
- Inception Labs
- Type
- Frontier model
- Open weights
- No
- Released
- 2026-03-04 (source: openrouter)
- Cheapest API price
- $0.250 in / $0.750 out per million tokens (via inception, observed 2026-10-13T20:47Z)
- Aggregate rank
- #614 of 743 (score 44.9)
Relative to other models, Mercury 2 is strongest on WeirdML (#48 of 109) and weakest on WebDev Arena (#65 of 67).
Across the 5 benchmarks where at least ten models are scored, Mercury 2 is in the top 10% on 0 and in the top half on 2.
Its cheapest listed input price, $0.250 per million tokens, is lower than that of 64% of the 427 models with a listed price.
Mercury 2 benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| WeirdML | 43.2% | #48 of 109 | Epoch AI Benchmarking Hub | 2026-10-13 |
| ALE-Bench | 786 | #36 of 79 | Epoch AI Benchmarking Hub | 2026-10-13 |
| CritPt | 0.8% | #72 of 115 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SciCode | 38.7% | #71 of 106 | Epoch AI Benchmarking Hub | 2026-10-13 |
| WebDev Arena | 1,167 | #65 of 67 | Epoch AI Benchmarking Hub | 2026-10-13 |
About Mercury 2
Mercury 2 is a frontier model from Inception Labs. It was released on 2026-03-04, according to the date OpenRouter first listed it (approximate). The cheapest API listing we track charges $0.250 per million input tokens and $0.750 per million output tokens, through inception. We hold 5 benchmark results for it across 3 areas, and none is in the top tenth of its board.
Where Mercury 2 stands
On our aggregate score, Mercury 2 is #614 of 743 ranked models, in the lower half. Its score is 44.9; the scale is centred on 50, so it sits 5.1 points below the typical model. That beats 17% of the ranked field. Among the 2 ranked Inception Labs models it is #2. The next one up is Mercury 2.5 (high) at 47.4.
Mercury 2, benchmark by benchmark
WeirdML puts Mercury 2 at 43.2%, #48 of 109. That beats 56% of the models tested, but 50.1 points short of GPT-6 Astra (max) (93.3%). The results either side are nemotron-3-ultra (43.5%) and grok-4-fast (42.9%). It covers unusual machine-learning tasks. The board median is 41.2%, so Mercury 2 is 2.0 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On ALE-Bench, Mercury 2 reaches 786. It ranks #36 of 79, ahead of 54% of the field, and 2,166 points behind GPT-6 Astra (max) at 2,951. The results either side are GPT-5 mini (high) (800) and gemini-2.5-pro_32K (786). It covers AtCoder heuristic-optimisation contests (performance rating). The board median is 764, so Mercury 2 is 22 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Mercury 2 lands below the middle on CritPt: 0.8%, #72 of 115, with 62% of the field scoring higher. GPT-5.6 Sol (max) leads at 32.3%; the closest result above is Trinity Large Thinking at 0.9%. It covers unpublished research-level physics problems. The board median is 2.9%, so Mercury 2 is 2.0 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Mercury 2 lands below the middle on SciCode: 38.7%, #71 of 106, with 66% of the field scoring higher. Claude Opus 5.5 (max) leads at 66.9%; the closest result above is GPT-5 mini (high) at 39.0%. It covers scientific research code. The board median is 44.3%, so Mercury 2 is 5.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
WebDev Arena is a weak spot. Mercury 2 scores 1,167, #65 of 67, behind 96% of the models measured. The leader, Claude Opus 5.5 (max), reaches 1,820. It covers building web apps, crowd-voted (Elo). The board median is 1,433, so Mercury 2 is 267 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Software engineering: uneven, ahead of 30% of the field on average across 3 benchmarks. Its best board here is ALE-Bench (#36 of 79) and its weakest is WebDev Arena (#65 of 67).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so Mercury 2 costs less than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen2.5-Coder-3B-Instruct ($0.010, 46.5). Similar scores cost more on DeepSeek-V3.2 (Thinking; Novita) ($0.259) and gemini-2.5-flash ($0.300).
Release history
The next one was Mercury 2.5 (high) on 2026-09-08, at 47.4.
Reading these numbers
The 5 scores on this page come from 1 source: Epoch AI Benchmarking Hub (5). No official leaderboard supplied any of them, so treat them as third-party measurements. With 5 results the aggregate rank can move a lot when one new score arrives.
Models ranked nearby
- #611 qwen2.5-32b-instruct (45.0)
- #612 Qwen3-30B-A3B (44.9)
- #613 aya-expanse-8b (44.9)
- #615 Qwen3.6 27B (44.9)
- #616 Nova 2.0 Pro Preview (unknown) (44.9)
- #617 internlm2-math-plus-7b (44.9)
More from Inception Labs
- Mercury 2.5 (high) (#439, 47.4)
Compare it on the live leaderboard → · How the aggregate score works