Inkling Small (xhigh): benchmark scores, price and ranking
Inkling Small (xhigh) from Thinking Machines ranks #244 of 743 LLMs on the LLMs Tiger aggregate benchmark score (49.7), based on 8 benchmark results captured up to 2026-10-13.
- Organisation
- Thinking Machines
- Type
- Open-weight / local model
- Open weights
- Yes
- Released
- 2026-07-15 (source: epoch)
- Cheapest API price
- $0.450 in / $1.20 out per million tokens (via deepinfra, observed 2026-10-13T19:47Z)
- Aggregate rank
- #244 of 743 (score 49.7)
Relative to other models, Inkling Small (xhigh) is strongest on GPQA Diamond (#43 of 275) and weakest on Mystery Games (#54 of 59).
Across the 8 benchmarks where at least ten models are scored, Inkling Small (xhigh) is in the top 10% on 0 and in the top half on 4.
Its cheapest listed input price, $0.450 per million tokens, is lower than that of 51% of the 427 models with a listed price.
Inkling Small (xhigh) benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| GPQA Diamond | 88.5% | #43 of 275 | Epoch AI Benchmarking Hub | 2026-10-13 |
| OTIS Mock AIME | 90.0% | #43 of 189 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Chess Puzzles | 18.0% | #51 of 122 | Epoch AI Benchmarking Hub | 2026-10-13 |
| ARC-AGI-2 | 40.1% | #40 of 87 | ARC Prize official leaderboard | 2026-10-13 |
| FrontierMath T1-3 | 46.3% | #40 of 68 | Epoch AI Benchmarking Hub | 2026-10-13 |
| FrontierMath T4 | 17.1% | #39 of 50 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SimpleQA Verified | 19.1% | #58 of 69 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Mystery Games | 6.0% | #54 of 59 | Epoch AI Benchmarking Hub | 2026-10-13 |
About Inkling Small (xhigh)
Inkling Small (xhigh) is an open-weight model that can run on your own hardware from Thinking Machines. It was released on 2026-07-15, according to Epoch AI. The cheapest API listing we track charges $0.450 per million input tokens and $1.20 per million output tokens, through deepinfra. We hold 8 benchmark results for it across 2 areas, and none is in the top tenth of its board.
Where Inkling Small (xhigh) stands
On our aggregate score, Inkling Small (xhigh) is #244 of 743 ranked models, around the middle of the table. Its score is 49.7; the scale is centred on 50, so it sits 0.3 points below the typical model. That beats 67% of the ranked field. Among the 2 ranked Thinking Machines models it is #1. The next one down is Inkling (xhigh) at 48.9.
Inkling Small (xhigh), benchmark by benchmark
GPQA Diamond puts Inkling Small (xhigh) at 88.5%, #43 of 275. That beats 84% of the models tested, but 7.3 points short of GPT-6 Astra (max) (95.8%). The results either side are grok-4.3 (high) (88.8%) and Darwin-36B-Opus (88.4%). It covers expert-level science questions. The board median is 71.7%, so Inkling Small (xhigh) is 16.8 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On OTIS Mock AIME, Inkling Small (xhigh) reaches 90.0%. It ranks #43 of 189, ahead of 77% of the field, and 10.0 points behind Claude Fable 5.1 (max) at 100.0%. The results either side are Qwen3.6 27B (91.1%) and Muse Spark (88.9%). It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so Inkling Small (xhigh) is 32.2 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Inkling Small (xhigh) manages 18.0% on Chess Puzzles, good for #51 of 122. GPT-6 Astra (max) leads at 72.0%, 54.0 points further up, and 58% of models sit below. The results either side are Qwen3.7 Max (19.0%) and o3-mini (high) (17.0%). It covers chess tactics puzzles. The board median is 12.5%, so Inkling Small (xhigh) is 5.5 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
ARC-AGI-2 puts Inkling Small (xhigh) at 40.1%, #40 of 87. That beats 54% of the models tested, but 54.9 points short of GPT-6 Astra (max) (95.0%). The results either side are Qwen3.8-27B (XHigh) (42.4%) and Opus 4.5 (Thinking, 64K) (37.6%). It covers abstract reasoning, ARC Prize public eval set. The board median is 22.8%, so Inkling Small (xhigh) is 17.4 points above it. Source: ARC Prize official leaderboard, captured 2026-10-13.
Inkling Small (xhigh) lands below the middle on FrontierMath T1-3: 46.3%, #40 of 68, with 57% of the field scoring higher. GPT-6 Astra (max) leads at 93.7%; the closest result above is GPT-5 mini (high) at 46.7%. It covers research-level maths, tiers 1-3. The board median is 55.6%, so Inkling Small (xhigh) is 9.3 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Inkling Small (xhigh) lands below the middle on FrontierMath T4: 17.1%, #39 of 50, with 76% of the field scoring higher. GPT-6.1 Sol (max) leads at 100.0%; the closest result above is grok-4.20-0309-reasoning at 17.1%. It covers research-level maths, tier 4. The board median is 30.5%, so Inkling Small (xhigh) is 13.4 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
SimpleQA Verified is a weak spot. Inkling Small (xhigh) scores 19.1%, #58 of 69, behind 83% of the models measured. The leader, gemini-4-argon-high, reaches 76.0%. It covers short factual questions. The board median is 41.0%, so Inkling Small (xhigh) is 21.9 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Mystery Games is a weak spot. Inkling Small (xhigh) scores 6.0%, #54 of 59, behind 90% of the models measured. The leader, GPT-6 Astra (max), reaches 84.0%. It covers deductive reasoning puzzles. The board median is 23.0%, so Inkling Small (xhigh) is 17.0 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Reasoning, maths and science: solid, ahead of 56% of the field on average across 5 benchmarks. Its best board here is GPQA Diamond (#43 of 275) and its weakest is FrontierMath T4 (#39 of 50). General knowledge and chat: weak, ahead of 28% of the field on average across 3 benchmarks. Its best board here is Chess Puzzles (#51 of 122) and its weakest is Mystery Games (#54 of 59).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so Inkling Small (xhigh) 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), Qwen3.7 Flash ($0.030, 50.5). Similar scores cost more on SWE-1.7 ($0.500) and Kimi-K2-Instruct-0905 ($0.500).
Release history
The next one was Inkling (xhigh) on 2026-07-15, at 48.9.
Reading these numbers
The 8 scores on this page come from 2 sources: Epoch AI Benchmarking Hub (7), ARC Prize official leaderboard (1). 1 of them come from an official leaderboard and 7 from datasets or third-party runs. The leaderboard also lists 6 other reasoning-effort settings of this model, which redirect here. This page shows the highest.
Inkling Small (xhigh) is also listed at 6 other reasoning-effort settings; this page shows the highest-effort row, which is the one the leaderboard keeps by default.
Models ranked nearby
- #241 Frogboss 32B 2510 (49.7)
- #242 claude-sonnet-4-20250514_16K (49.7)
- #243 Qwen3-235B-A22B-Thinking-2507 (49.7)
- #245 GLM-4.5-FP8 (49.6)
- #246 GLM-5.1 (49.6)
- #247 Opus 4.5 (Thinking, 64K) (49.6)
More from Thinking Machines
- Inkling (xhigh) (#314, 48.9)
Compare it on the live leaderboard → · How the aggregate score works