Inference
EN
Inception

Mercury 2.5

Mercury 2.5 is a reasoning model from Inception. 17 benchmarks count toward its index, in 6 categories.

  • Closed weights
  • Reasoning
  • Input: text
  • Released 2026-09
  • Retires 8 Nov 2026
RankedEnough results for an index.
IndexAll categories, out of 100.46.2±5.4Median 51.0
CoverageIndex weight that has results.85%
SpeedOutput tokens per second.125/sMedian 67/s
First tokenSeconds to the first output token.0.7 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.04Median $0.60
Output / 1MUS dollars per 1M output tokens.$0.15Median $2.55
ContextMaximum tokens in one request.260KMedian 262K
EloLMArena rating and rank.—Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
inception/mercury-2.5
Max outputMaximum output tokens per request.
66K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
<$0.01
WeightsLink to the published weights.
—
RetiresDate OpenRouter removes the model.
8 Nov 2026

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.5 counted49.3133 / 218
CodingCode writing and repair.5 counted44.5165 / 209
ReasoningLogic problems and puzzles.2 counted45.4142 / 227
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.3 counted41.4186 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted61.0—
MathMath problems.1 counted48.2—

Results

17 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
τ³-Bench Tool-Agent-User EvaluationAgenticAbove fitFar above this model's index.96.0%1 / 32
Instruction Following BenchmarkInstruction77.0%22 / 57
Artificial Analysis Long Context ReasoningReasoning68.0%125 / 217
Artificial Analysis Humanity's Last ExamKnowledge11.8%137 / 222
Scientific Code BenchmarkCoding38.0%24 / 35
Critical Physics TasksReasoning0.0%150 / 214
Artificial Analysis Omniscience AccuracyKnowledge22.0%120 / 164
GPQA Diamond · BenchLMKnowledge79.0%63 / 84
GDPval-AA normalizedAgentic0.0%101 / 133
Artificial Analysis SciCodeCoding38.5%95 / 114
Vibe Code Bench v1.1Coding3.7% ±1.4111 / 123
Terminal-Bench 2.1 · Vals AIAgentic34.5% ±0.770 / 75
Code MigrationCoding4.5% ±1.073 / 77
ProofBench v1.1Math3.0% ±1.752 / 54
Terminal-Bench 4.0Agentic0.0%59 / 61
DeepSearchQAAgentic34.0%24 / 24
SkillsBenchCoding18.1% ±3.135 / 35
IOICoding2.4% ±0.745 / 45
LegalBenchLegal83.1% ±0.558 / 149
CyberBench v1.1Other45.5% ±6.141 / 48
MedCodeHealthcare31.3% ±2.094 / 106
Terminal-Bench-ScienceScience0.0%41 / 44
Public Benefits Bench v1.1Public benefits45.3% ±1.347 / 50
Legal Research BenchLegal4.3% ±1.485 / 90
Finance Agent (v2)Finance18.9% ±0.276 / 79
MedScribeHealthcare55.1% ±2.1106 / 109
EMBFinance8.8% ±1.374 / 75
Tax Agent BenchFinance2.5% ±1.082 / 82

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseVals AINo licence stated

History

5 releases
Index46.6 → 46.2
23 Oct 202612 Nov 2026
Output / 1M$0.15 → $0.15
23 Oct 202612 Nov 2026

Questions

How much does Mercury 2.5 cost?
$0.04 per 1M input tokens, $0.15 per 1M output tokens.
How good is Mercury 2.5?
Its index is 46.2 out of 100. That is place 175 of 283 ranked models.
What is Mercury 2.5 best at?
Agentic: 49.3 out of 100, place 133 of 218.
How fast is Mercury 2.5?
125 output tokens per second, 0.7 s to the first token.
What is the context window of Mercury 2.5?
260K tokens.
Is Mercury 2.5 open source?
No. Its weights are not published.
When was Mercury 2.5 released?
September 2026.

Badge

Mercury 2.5 on Inference360: place 175 of 283, index 46.2
HTML
<a href="https://inference360.ai/models/mercury-2-5"><img src="https://inference360.ai/badges/mercury-2-5.svg" alt="Mercury 2.5 on Inference360: place 175 of 283, index 46.2" height="20"></a>
Markdown
[![Mercury 2.5 on Inference360: place 175 of 283, index 46.2](https://inference360.ai/badges/mercury-2-5.svg)](https://inference360.ai/models/mercury-2-5)

Compare with

Nemotron 3.5 Bolt 74B-A9B46.4Gemini 2.5 Flash (Preview)46.6Grok 4 Fast (Reasoning)45.3Mercury 352.3

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from Inception

All models
Mercury 2.5 PreviewUnrankedMercury 352.3Mercury Edit 2UnrankedMercury 2UnrankedMercury DecideUnranked