Inkling released
Thinking Machines Lab · Model release
Source confirmed
Data as of October 14, 2026 · How the score is built
CurrentOpen WeightHybrid
Released Jul 15, 20261M context
Thinking Machines Inkling model
Decision readingInkling scores 52.1 out of 100 and ranks #89 of 221. This profile shows 24 source-displayable benchmark rows; its strongest eligible category is Mathematics at #11. API pricing is $1.87 input and $4.68 output per million tokens, with cached input at $0.374.
Released Jul 15, 2026 — see all recent releases
Inkling will be repriced, updated or retired. Get each notice with its source and date. Follow model changes
Decision snapshot
Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.
Capability
52.1/100
field median 49.8#89 of 221 ranked models
Public
#89of 221
Verified #51 of 85
Price
$1.87input / $4.68 output
input median $0.80cached $0.37 · blended $3.27
Speed
187tok/s
field median 90 tok/sFirst token 12.49 s
Context
1Mtokens
field median 256,000Reported for this model; direct source link not stored
Strongest published evidence
Mathematics ranks #11. Particularly strong for mathematical reasoning, scientific computing, and quantitative analysis.
Validate before choosing
24 published rows leave some tracked benchmark slots empty. Coding is its lowest eligible category at #80.
Source-linked · 24 displayable benchmark rows
Follow model changesCategory score record
Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticRank #72 of 124Percentile 42ndWeight 22%5 benchmarksVerified | 36.2 | 5 benchmarks | Verified | |||
| CodingRank #80 of 149Percentile 47thWeight 20%6 benchmarksVerified | 34.0 | 6 benchmarks | Verified | |||
| ReasoningWeight 17%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| MultimodalRank #35 of 54Percentile 36thWeight 12%3 benchmarksVerified | 67.5 | 3 benchmarks | Verified | |||
| KnowledgeRank #62 of 181Percentile 66thWeight 12%6 benchmarksVerified | 54.5 | 6 benchmarks | Verified | |||
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| Inst. FollowingRank #35 of 128Percentile 73rdWeight 5%1 benchmarkVerified | 87.6 | 1 benchmark | Verified | |||
| MathRank #11 of 12Percentile 9thWeight 5%3 benchmarksVerified | 58.3 | 3 benchmarks | Verified |
24 of 670 tracked benchmark slots have displayable evidence · bars run 0–100
Coverage detailsHow much of this is verified
Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.
- Agentic5/5 verified
- Coding6/6 verified
- ReasoningNot measured
- Multimodal3/3 verified
- Knowledge6/6 verified
- MultilingualNot measured
- Inst. Following1/1 verified
- Math3/3 verified
Verified sourceProvisionalNot measured
Capability shape
Each axis shows percentile within that category’s eligible cohort. The comparison outline is the median of the six nearest public-score peers; a collapsed vertex means the category is not rank-eligible.
Inkling category percentile values
- Agentic42nd percentile
- Coding47th percentile
- ReasoningNot eligible
- Multimodal36th percentile
- Knowledge66th percentile
- MultilingualNot eligible
- Instruction following73rd percentile
- Math9th percentile
The dashed outline is median of 6 nearest peers.
Eligible category ranks
- Agentic#72/124
- Coding#80/149
- ReasoningNot ranked
- Multimodal#35/54
- Knowledge#62/181
- MultilingualNot ranked
- Inst. Following#35/128
- Math#11/12
Top decileTop quartileMid-fieldNot eligible
Benchmark ledger
Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.
Coding6 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| SWE-bench Pro | Score54.3% | Versus best verified row Best verified: Claude Opus 5.5 · 89.9% | Gap35.6 behind | Weight26% ref. weight | Provider exact |
| FrontierSWE v2 | Score4.1% | Versus best verified row Best verified: GPT-6 Astra · 65.5% | Gap61.4 behind | Weight8% ref. weight | Benchmark exact |
| LiveCodeBench (Vals)LiveCodeBench, Vals AI run | Score85.5% | Versus best verified row Best verified: Claude Fable 5.1 · 90.5% | Gap5 behind | Weight8% ref. weight | |
| Terminal-Bench 2.1 | Score63.8% | Versus best verified row Best verified: SWE-2 · 92.8% | Gap29 behind | WeightScored in Agentic | Provider exact |
| SWE-bench VerifiedSoftware Engineering Benchmark Verified | Score77.6% | Versus best verified row Best verified: Claude Opus 5 · 96% | Gap18.4 behind | WeightDisplay only | Provider exact |
| SWE-bench (Vals)SWE-bench, Vals AI run | Score77.6% | Versus best verified row Best verified: Claude Opus 5 · 97.0% | Gap19.4 behind | WeightDisplay only | Verified |
Agentic5 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| Terminal-Bench 2.1 | Score63.8% | Versus best verified row Best verified: SWE-2 · 92.8% | Gap29 behind | Weight8% ref. weight | Provider exact |
| BrowseComp | Score77.1% | Versus best verified row Best verified: Atria Dawn Preview · 92.5% | Gap15.4 behind | Weight8% ref. weight | Provider exact |
| MCP Atlas | Score74.1% | Versus best verified row Best verified: Muse Spark 1.1 · 88.1% | Gap14 behind | Weight4% ref. weight | Provider exact |
| Terminal-Bench 2.1 (Vals)Terminal-Bench 2.1, Vals AI run | Score47.6% | Versus best verified row Best verified: GPT-6 Astra · 87.3% | Gap39.7 behind | Weight3% ref. weight | |
| CWE-bench v1 | Score37.0% | Versus best verified row Best verified: GPT-6 Astra · 68.0% | Gap31 behind | WeightDisplay only | Benchmark exact |
Multimodal3 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| MMMU-ProMassive Multi-discipline Multimodal Understanding Pro | Score73.5% | Versus best verified row Best verified: Gemini 3.5 Flash · 83.6% | Gap10.1 behind | WeightWeighted 40% | Provider exact |
| CharXivCharXiv Reasoning | Score82% | Versus best verified row Best verified: Qwen3.8 Max · 93.5% | Gap11.5 behind | WeightWeighted 20% | Provider exact |
| CharXiv w/o toolsCharXiv Reasoning without tools | Score78.1% | Versus best verified row Best verified: Claude Mythos 5 · 88.9% | Gap10.8 behind | WeightDisplay only | Provider exact |
Knowledge6 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| HLEHumanity's Last Exam | Score46% | Versus best verified row Best verified: Claude Fable 5.1 · 65% | Gap19 behind | Weight44% ref. weight | Provider exact |
| HLE w/o toolsHumanity's Last Exam without tools | Score30% | Versus best verified row Best verified: Claude Opus 5.5 · 64.4% | Gap34.4 behind | Weight7% ref. weight | Provider exact |
| MMLU-Pro (Vals)MMLU-Pro, Vals AI run | Score86.3% | Versus best verified row Best verified: Claude Fable 5.1 · 92.4% | Gap6.1 behind | Weight6% ref. weight | Verified |
| GPQAGraduate-Level Google-Proof Q&A | Score87.9% | Versus best verified row Best verified: GPT-6 Astra · 96% | Gap8.1 behind | Weight3% ref. weight | Provider exact |
| GPQA Diamond (Vals)GPQA Diamond, Vals AI run | Score87.1% | Versus best verified row Best verified: Gemini 3.1 Pro · 95.5% | Gap8.4 behind | Weight2% ref. weight | |
| GPQA-DGPQA Diamond | Score87.9% | Versus best verified row Best verified: GPT-6 Astra · 96.0% | Gap8.1 behind | WeightDisplay only | Provider exact |
Inst. Following1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| IFBenchInstruction Following Benchmark | Score79.8% | Versus best verified row Best verified: MAI-Thinking-1 · 85% | Gap5.2 behind | WeightWeighted 70% | Provider exact |
Math3 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| FrontierMath v2 (Tiers 1-3)FrontierMath v2 Tiers 1-3 | Score33.333% | Versus best verified row Best verified: GPT-6 Astra · 93.684% | Gap60.4 behind | WeightWeighted 30% | Benchmark exact |
| AIME26AIME 2026 | Score97.1% | Versus best verified row Best verified: GLM-5.2 · 99.2% | Gap2.1 behind | WeightWeighted 25% | Provider exact |
| FrontierMath v2 (Tier 4)FrontierMath v2 Tier 4 | Score4.878% | Versus best verified row Best verified: GPT-6.1 Sol · 100.000% | Gap95.1 behind | WeightWeighted 10% | Benchmark exact |
Bars run 0–100; the dark tick marks the best source-verified value
All 24 rowsWhat it costs to get this score
Published API price against the public score. The x-axis uses a log scale; the dashed path marks models that are not beaten by a cheaper, higher-scoring option. Price uses average of published input and output rates.
Current modelExplore all models
Inkling · 52.1 score · $3.27 blended per million tokens
The chart opens on the current model. Scroll horizontally to inspect the full price axis.
Horizontal: blended price per million tokens, log scale · Vertical: public score
Lineage
The sequence follows explicit supersedes links. Each score is estimated for that model; a relative can inform a sparse estimate but never sets a floor, so a newer release can score below an earlier one. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Base entry
Inkling-Small · 53.4Radar
Full release historyInkling release history
Radar confirmed these at the source. Use Inkling in your work? Explore Radar to follow supported changes and choose your alerts.
Radar
Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- Not published
- Context window
- 1M
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- Not sourced yet
- Output modalities
- Not sourced yet
- Parameters
- Not sourced yet
- Availability
- Not sourced yet
- Cloud regions
- Not tracked yet
- Lifecycle
- Current
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Published at $0.37 per million cached input tokens
- Self-host
- Open weights available; hardware estimate not sourced
- Rate limits
- Not tracked yet
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
Inkling ranks #89 of 221 on the public leaderboard with a score of 52.13/100. Its source-verified position is #51 of 85.
Inkling is a open weight model with a 1M context window. No explicit reasoning mode is documented in this profile.
Inkling sits in the Inkling family with Inkling-Small. 24 of 670 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.
Its strongest eligible category is Mathematics at #11, while its lowest eligible position is Coding at #80. particularly strong for mathematical reasoning, scientific computing, and quantitative analysis.
Last updated October 14, 2026. Runtime fields remain blank until a sourced snapshot exists.
Deployment options
Self-host and provider-specific paths stay separate from benchmark evidence so operating constraints are visible before a score becomes the whole decision.
Published weights are available, but BenchLM does not yet have a sourced parameter and VRAM profile for this exact model. Hardware cost estimates stay unavailable until that sizing record is complete.
Estimate VRAM from known parametersQuestions
How does Inkling perform overall in AI benchmarks?
Inkling ranks #89 out of 221 models on the public BenchAlign leaderboard, with a score of 52.13/100. Its evidence status is Supported, and this profile shows 24 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.
Is Inkling good for knowledge and understanding?
Inkling ranks #62 out of 181 eligible models for knowledge and understanding, with a public category score of 54.5/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Inkling good for coding and programming?
Inkling ranks #80 out of 149 eligible models for coding and programming, with a public category score of 34/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Inkling good for mathematics?
Inkling ranks #11 out of 12 eligible models for mathematics, with a public category score of 58.3/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Inkling good for agentic tool use and computer tasks?
Inkling ranks #72 out of 124 eligible models for agentic tool use and computer tasks, with a public category score of 36.2/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Inkling good for multimodal and grounded tasks?
Inkling ranks #35 out of 54 eligible models for multimodal and grounded tasks, with a public category score of 67.5/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Inkling good for instruction following?
Inkling ranks #35 out of 128 eligible models for instruction following, with a public category score of 87.6/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Inkling open source?
Inkling is an open-weight model from Thinking Machines Lab. Its weights can be downloaded for local or hosted deployment, subject to the published license. Open weight does not automatically mean open source: training data and training code may remain private, and commercial restrictions can still apply.
Which sibling models are related to Inkling?
Inkling belongs to the Inkling family. Related tracked variants include Inkling-Small. A sibling link indicates shared lineage or a documented configuration relationship; it does not mean the variants have identical pricing, context limits, benchmark evidence, or deployment behavior. Compare before switching.
Does Inkling have full benchmark coverage on BenchLM?
No. Inkling currently has 49 source-displayable rows across 670 tracked benchmark slots. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.
What is the context window size of Inkling?
Inkling has a reported context window of 1M in the exact-model catalog record. The value stays visible, but the profile marks its source link as unavailable instead of presenting it as directly documented. Maximum output length remains separate because providers often publish a different limit.