Thinking Machines Labopen-weight
Released: 2026-07-15Artificial AnalysisInkling
Inkling from Thinking Machines, tracked on the Artificial Analysis Intelligence Index.
Intelligence Index
12.5/ 100
Artificial Analysis Intelligence Index (independent composite)Throughput Speed
80tok/sec
Output streaming throughputTime to First Answer
21s
Time to first answer token, including any thinkingCost to Run the Index
1.35USD
Artificial Analysis cost of the full Index runMeasured results · Artificial Analysis
Independent evaluations of Inkling (Xhigh configuration). Values are reported as published; nothing here is estimated or derived. Snapshot 2026-11-15.
- Intelligence Index
- 12.5
- Cost to run the Index
- $1.35
- Time per task
- 10.7 min
- Omniscience index
- -1.5
- Reasoning model
- Yes
- Cached-input price
- $0.17 / 1M
Reasoning & knowledge
- AA-Omniscience Accuracy39.3%
- CritPt5.4%
- CritPt v2 (Hard)18.0%
- GPQA Diamond87.2%
- HLE-Diamond10.6%
- Humanity's Last Exam31.9%
Coding & terminal agents
- FrontierCode 1.115.8%
- FrontierCode-AA 2.03.9%
- SciCode47.0%
- Terminal-Bench 2.155.1%
- Terminal-Bench 4.01.0%
- Terminal-Bench-Science 0.10.0%
Agentic work
- AA-AnalystAgent23.8%
- AA-Briefcase v1.1 (Elo)833 Elo
- AutomationBench-AA5.0%
- EnterpriseOps-Gym-AA38.0%
- GDPval-AA (normalized Elo)1080 Elo
- Harvey LAB-AA0.0%
- τ³-Banking29.1%
Long context & documents
Multimodal & science
- AA-3D (Elo)591 Elo
- Code2Video-AA (Elo)467 Elo
- MMMU-Pro73.5%
Health & safety
Headline Evaluation Breakdown
Six independently measured evaluations (Artificial Analysis)Reasoning · CritPt v2 (Hard)18.0%
Coding & agents · Terminal-Bench 4.01.0%
Long context · AA-LCR77.3%
Knowledge · Humanity's Last Exam31.9%
Sourced Benchmark Evaluations (22)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| CritPt v2 (Hard) | Active | 18% | 93 | 2026-11-15 | independent | Source ↗ |
| EnterpriseOps-Gym-AA | Active | 38% | 93 | 2026-11-15 | independent | Source ↗ |
| Harvey LAB-AA | Active | 0% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-AnalystAgent | Active | 23.8% | 93 | 2026-11-15 | independent | Source ↗ |
| AutomationBench-AA | Active | 5% | 93 | 2026-11-15 | independent | Source ↗ |
| MLCR-AA (Medical Long Context Reasoning) | Active | 6.1% | 93 | 2026-11-15 | independent | Source ↗ |
| GDP.pdf | Active | 12.8% | 93 | 2026-11-15 | independent | Source ↗ |
| FrontierCode-AA 2.0 | Active | 3.9% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-Omniscience (Non-Hallucination) | Active | 32.8% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-Omniscience (Accuracy) | Active | 39.3% | 93 | 2026-11-15 | independent | Source ↗ |
| HLE-Diamond | Active | 10.6% | 93 | 2026-11-15 | independent | Source ↗ |
| FrontierCode | Active | 15.8% | 98 | 2026-11-15 | independent | Source ↗ |
| CritPt | Active | 5.4% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-LCR (Long Context Reasoning) | Active | 77.3% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench-Science 0.1 | Active | 0% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench 4.0 | Active | 1% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench 2.1 | Active | 55.1% | 93 | 2026-11-15 | independent | Source ↗ |
| τ³-Banking | Active | 29.1% | 93 | 2026-11-15 | independent | Source ↗ |
| SciCode | Active | 47% | 93 | 2026-11-15 | independent | Source ↗ |
| MMMU-Pro | Active | 73.5% | 93 | 2026-11-15 | independent | Source ↗ |
| Humanity's Last Exam | Active | 31.9% | 93 | 2026-11-15 | independent | Source ↗ |
| GPQA | Nearing Saturation | 87.2% | 78 | 2026-11-15 | independent | Source ↗ |
Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run thinking-machines-inkling
# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "thinking-machines-inkling",
"messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
}'Evaluation & Sourcing Notes
Scores listed for Inkling represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.
All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.