# Inkling 2 Mini Benchmark Scores & Performance

> Inkling 2 Mini by Thinking Machines Lab scores 55.9/100 overall, ranking #70 out of 1034 AI models.

Canonical page: https://benchlm.ai/models/inkling-2-mini

Last updated: November 21, 2026

## Model Details

| Property | Value |
|----------|-------|
| Creator | Thinking Machines Lab |
| Source Type | Open Weight |
| Reasoning Type | Hybrid |
| Context Window | 1M |
| Official model card | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| Overall Score | 55.9/100 |
| Overall Rank | #70 of 1034 |

## Family & Coverage

- Family: Inkling
- Variant: mini (2 Mini)
- Benchmarks covered: 50 of 828
- Sibling models: [Inkling 2 Small](/models/inkling-2-small), [Inkling](/models/inkling), [Inkling-Small](/models/inkling-small)
- Coverage note: BenchLM currently has partial benchmark coverage for this model, so the overall score is conservative.

## Agentic Benchmarks

| Benchmark | Score | Who ran it | Source |
|-----------|-------|------------|--------|
| [Terminal-Bench 4.0](/benchmarks/terminal-bench) | 5.60% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [Terminal-Bench 2.1](/benchmarks/terminalbench21) | 71.6% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [Terminal-Bench-Science 0.1](/benchmarks/terminal-bench-science) | 1.4% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [BrowseComp](/benchmarks/browsecomp) | 79% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [DeepSearchQA](/benchmarks/deepsearchqa) | 85.3% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [MCP Atlas](/benchmarks/mcpatlas) | 82.7% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [JobBench](/benchmarks/jobbench) | 49.6% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [EnterpriseOps-Gym](/benchmarks/enterpriseopsgym) | 42.00% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [AA-AnalystAgent](/benchmarks/aaanalystagent) | 46.3% | Carried from an upstream page | [Artificial Analysis: aa-analyst-agent leaderboard](https://artificialanalysis.ai/evaluations/aa-analyst-agent) |

## Coding Benchmarks

| Benchmark | Score | Who ran it | Source |
|-----------|-------|------------|--------|
| [FrontierCode 2.0](/benchmarks/frontiercode-2) | 15.1% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [DeepSWE](/benchmarks/deepswe) | 48.2% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [Terminal-Bench 2.1](/benchmarks/terminalbench21) | 71.6% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [SWE Multilingual](/benchmarks/swe-bench-multilingual) | 73.7% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [SWE-bench Verified](/benchmarks/swe-bench-verified) | 79.2% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |

## Multimodal & Grounded Benchmarks

| Benchmark | Score | Who ran it | Source |
|-----------|-------|------------|--------|
| [PerceptionBench](/benchmarks/perceptionbench) | 57.5% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [MMMU-Pro](/benchmarks/mmmu-pro) | 74.9% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [OmniDocBench 1.6](/benchmarks/omnidocbench16) | 92.8% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [BabyVision](/benchmarks/babyvision) | 53.4% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [BabyVision w/ Python](/benchmarks/babyvisionpython) | 88.8% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [ScreenSpot Pro](/benchmarks/screenspot-pro) | 76.5% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [CharXiv](/benchmarks/charxiv) | 88.3% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [CharXiv w/o tools](/benchmarks/charxivnotools) | 80.6% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [Chartography (no tools)](/benchmarks/chartography) | 22.1% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [Chartography (tools)](/benchmarks/chartographywithtools) | 72.5% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [ZeroBench](/benchmarks/zerobench) | 17.0% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [ZeroBench w/ Python](/benchmarks/zerobenchpython) | 34.0% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [MathVision](/benchmarks/mathvision) | 89.0% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |

## Reasoning Benchmarks

| Benchmark | Score | Who ran it | Source |
|-----------|-------|------------|--------|
| [LongBench v2](/benchmarks/longbench-v2) | 66.9% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [MRCR v2 256K-512K](/benchmarks/mrcr-v2-256k-512k) | 46.4% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |

## Knowledge Benchmarks

| Benchmark | Score | Who ran it | Source |
|-----------|-------|------------|--------|
| [GPQA](/benchmarks/gpqa) | 86.6% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [GPQA-D](/benchmarks/gpqa-diamond) | 86.6% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [HLE w/o tools](/benchmarks/hlenotools) | 30.7% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [Artificial Analysis Intelligence Index](/benchmarks/artificialanalysis) | 20.1% | Carried from an upstream page | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/inkling-2-mini) |
| [AA-Omniscience Index](/benchmarks/aaomniscienceindex) | -8.6% | Carried from an upstream page | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/inkling-2-mini) |
| [AA-Omniscience Accuracy](/benchmarks/omniscienceaccuracy) | 20.5% | Independent run | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/inkling-2-mini) |
| [AA-Omniscience Hallucination Rate](/benchmarks/omnisciencehallucinationrate) | 36.6% | Carried from an upstream page | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/inkling-2-mini) |

## Instruction Following Benchmarks

| Benchmark | Score | Who ran it | Source |
|-----------|-------|------------|--------|
| [MultiChallenge (provider reports)](/benchmarks/multichallenge-provider-reports) | 50.00% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |

## Mathematics Benchmarks

| Benchmark | Score | Who ran it | Source |
|-----------|-------|------------|--------|
| [Apex](/benchmarks/apex) | 34.4% | Reported by Thinking Machines Lab | [Thinking Machines Lab: Introducing Inkling 2 Mini](https://thinkingmachines.ai/news/inkling-2-mini/) |

## Lab-reported results

Thinking Machines Lab published these results with a metric, split or setting no scored field uses, or for a test with no BenchLM page. They are shown for reference and never count toward a rank.

| Benchmark | Reported value | What was measured | Who ran it | Use in ranking | Source |
|-----------|----------------|-------------------|------------|----------------|--------|
| [SAGE v1.1](/benchmarks/sage) | 85.9% | Score · Reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [ARC-AGI-1](/benchmarks/arcagi1) | 84.5% | Accuracy · Reported in the Inkling 2 Mini launch table and model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [AA-LCR v1.1](/benchmarks/lcr) | 80.0% | Score · Reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [SWE-bench Pro v2 (Hard)](/benchmarks/scale-swe-bench-pro-public-v2-hard) | 77.9% | Tasks resolved · 51-task Hard split using mini-swe-agent v2.4.6. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [Vibe Code Bench v1.1](/benchmarks/vibe-code-bench-1-100) | 77.9% | Score · openhands (vals); reported in the Inkling 2 Mini launch table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [MuirBench](/benchmarks/muirbench) | 77.5% | Overall accuracy · Reported in the Inkling 2 Mini launch table; not part of the original MuirBench ICLR 2025 paper table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [Excel Modeling Benchmark v1.1](/benchmarks/emb) | 66.0% | Score · Reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [AA-Omniscience v1.1 (non-hallucination)](/benchmarks/omnisciencehallucinationrate) | 63.4% | Non-hallucination rate · Reported in the Inkling 2 Mini launch table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [Finance Agent v2.1](/benchmarks/finance-agent-v2-1) | 56.5% | Score · Reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [GroundingME](/benchmarks/groundingme) | 55.5% | Overall accuracy · Reported in the Inkling 2 Mini model card; not yet evaluated on the official GroundingME CVPR 2026 leaderboard. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/news/inkling-2-mini/model-card/) |
| [Vals Finance Index](/benchmarks/vals-finance-index) | 50.1% | Index score · Reported in the Inkling 2 Mini launch table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [ARC-AGI-2](/benchmarks/arc-agi-2) | 49.3% | Accuracy · Reported in the Inkling 2 Mini launch table and model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [Vals Index (v2.2)](/benchmarks/valsindex) | 46.5% | Overall score · Reported in the Inkling 2 Mini launch table under Third Party Index. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [AA-AnalystAgent](/benchmarks/aaanalystagent) | 46.3% | Score · Reported in the Inkling 2 Mini launch table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [AutomationBench-AA v1.0.6](/benchmarks/aaautomationbench) | 41.8% | Task completion · Reported in the Inkling 2 Mini launch table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [SWE-bench Multimodal v2](/benchmarks/valsswebenchmultimodal) | 37.9% | Tasks resolved · mini-swe-agent (vals); reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [Code Migration v1.1](/benchmarks/codemigration) | 23.2% | Score · pi (vals); reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [MysteryMechanism](/benchmarks/mysterymechanism) | 23.0% | Score · Reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [ProofBench v1.1](/benchmarks/valsproofbench) | 18.0% | Score · Reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [GDP.pdf](/benchmarks/aagdppdf) | 16.8% | Score · Reported in the Inkling 2 Mini launch table under Vision Interpretation and Creation. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [HLE-Diamond](/benchmarks/hle-diamond) | 15.2% | Accuracy · Reported in the Inkling 2 Mini launch table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [Tax Agent Bench](/benchmarks/valstaxagentbench) | 14.1% | Score · Reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [HoneyBench v0.2](/benchmarks/honeybench) | 56.7% | Reward-hacking rate (lower is better) · Reported in the Inkling 2 Mini model card before inclusion on Goodhart Labs' HoneyBench v0.2 leaderboard. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/news/inkling-2-mini/model-card/) |
| [GDPVal-AA v2.1](/benchmarks/gdpvalaa) | 1319 | Elo rating · Reported in the Inkling 2 Mini launch table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [AA-Briefcase v1.1](/benchmarks/aabriefcaseelo) | 1263 | Elo rating · Reported in the Inkling 2 Mini launch table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [Code2Video-AA](/benchmarks/code2video) | 870 | Rating · Reported in the Inkling 2 Mini model card. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini model card](https://thinkingmachines.ai/model-card/inkling-2-mini/) |
| [AA Index (v5.0)](/benchmarks/artificialanalysis) | 20.1 | Intelligence Index · Reported in the Inkling 2 Mini launch table under Third Party Index. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |
| [AA-Omniscience v1.1](/benchmarks/aaomniscienceindex) | -8.6 | Omniscience score · Reported in the Inkling 2 Mini launch table. | Reported by Thinking Machines Lab | Shown only | [Thinking Machines Lab Inkling 2 Mini launch post](https://thinkingmachines.ai/news/inkling-2-mini/) |

## Other Thinking Machines Lab Models

- [Inkling 2 Small](/models/inkling-2-small) - Score: 66.7
- [Inkling](/models/inkling) - Score: 48.9
- [Inkling-Small](/models/inkling-small) - Score: 48.9
