What is the IQ of Inkling-Small?

Inkling-Small has an estimated IQ of 110, which ranks #81 out of 176 public models on the AI IQ leaderboard. The score combines seven equally weighted reasoning dimensions using measured benchmark results and conservative estimates for missing coverage.

What is Inkling-Small best at?

Among the dimensions with published benchmark results, Inkling-Small ranks highest in Academic Reasoning (IQ 125, #74 of 203) and Knowledge Work (IQ 114, #65 of 176). Its lowest-ranked measured dimension is Reliability (IQ 106, #124 of 177).

What is the EQ of Inkling-Small?

Inkling-Small doesn't have an EQ score yet. EQ (emotional reasoning) is calculated from EQ-Bench 4 and AttuneBench, and neither has published a result for Inkling-Small. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.

Which benchmarks has Inkling-Small been tested on?

Inkling-Small has published results on 27 of the 39 benchmarks behind AI IQ's Composite IQ, including LiveCodeBench, GPQA Diamond, MCP Atlas, SciCode and CritPt.

How much does Inkling-Small cost?

Inkling-Small costs $0.30 per million input tokens and $1.20 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $1.24 per million input-plus-output tokens.

Is Inkling-Small open source?

Yes — Inkling-Small is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.

Who developed Inkling-Small?

Inkling-Small was developed by Thinking Machines, based in the United States.

When was Inkling-Small released?

Inkling-Small was publicly released on July 30, 2026.

How large is the context window of Inkling-Small?

Inkling-Small supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.

How fast is Inkling-Small?

In independent measurements, Inkling-Small generates about 224 tokens per second, with a typical response time of about 13 seconds.

Thinking MachinesReleased Jul 30, 2026

Open weights

Inkling-Small

Its best results are in Academic Reasoning (#74 of 203) and Knowledge Work (#65 of 176); weakest in Reliability (#124 of 177). All seven dimensions are measured.

Composite IQ ?

110

#81 of 176 ranked models · ahead of 54%

US (Other)'s #2 model, behind Beam (114)

Effective cost

$1.24/1M

$0.30 in · $1.20 out

Speed

224tok/s

13s typical response

Context

1Mtokens

1,000,000 tokens in

Evidence

7of 7 dims

27 of 39 benchmarks

Compare with gpt-6.1-solAPI pricing

IQ profile

Seven dimensions, averaged equally into Composite IQ

Compare shape with

Measured95% range

#1 in dimensionTop-10 median

IQ dimensions

Ranks are among models scored in each dimension. Ranges are model-based estimates, not calibrated confidence intervals.

DimensionIQPosition · 70–160RankEvidencevs Top 10

Abstract ReasoningAbstract

91measured

#99 / 201

4/5

-35.5

Mathematical ReasoningMath

115measured

#72 / 176

3/5

-32

Academic ReasoningAcademic

125measured

#74 / 203

5/6

-10.5

Programmatic ReasoningCoding

107measured

#103 / 177

4/6

-42.5

Computer UseComputer Use

114measured

#75 / 176

3/5

-29

ReliabilityReliability

106measured

#124 / 177

3/6

-28

Knowledge WorkKnowledge Work

114measured

#65 / 176

5/6

-17.5

Inkling-Small95% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human

Benchmark results

27 of 39 Composite IQ benchmarks have a published result for Inkling-Small. Benchmarks without a result stay listed so gaps are visible.

Abstract Reasoning IQ 91 · measuredBest: #39 of 68 on ARC-AGI-24/5

BenchmarkScorePercentileRankLeader · source

ARC-AGI-2

Full source leaderboard reconciliation · Oct 23

40.1%

#39 / 68

Leader gpt-6-astra · 95.0%
Full source leaderboard reconciliation · Oct 23

ARC-AGI-1

Full source leaderboard reconciliation · Oct 23

84.0%

#40 / 68

Leader fable-5 · 98.5%
Full source leaderboard reconciliation · Oct 23

Chess Puzzles

Epoch AI vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 15

18.0%

#57 / 76

Leader gpt-6-astra · 72.0%
Epoch AI vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 15

Mystery Game Puzzles

Full source leaderboard reconciliation vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 23

6.0%

#54 / 55

Leader gpt-6-astra · 84.0%
Full source leaderboard reconciliation vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 23

Not yet reported

ARC-AGI-3

Mathematical Reasoning IQ 115 · measuredBest: #50 of 84 on FrontierMath Tier 1-33/5

BenchmarkScorePercentileRankLeader · source

FrontierMath Tier 1-3

Full source leaderboard reconciliation · Oct 23

46.3%

#50 / 84

Leader gpt-6-astra · 93.7%
Full source leaderboard reconciliation · Oct 23

FrontierMath Tier 4

Full source leaderboard reconciliation · Oct 23

17.1%

#50 / 72

Leader gpt-6.1-sol · 100%
Full source leaderboard reconciliation · Oct 23

ProofBench 1.1

Full source leaderboard reconciliation v1.1 · Oct 27

6.0%

#46 / 48

Leader fable-5.1 · 100%
Full source leaderboard reconciliation v1.1 · Oct 27

Not yet reported

MathArenaAIME

Academic Reasoning IQ 125 · measuredBest: #53 of 147 on GPQA Diamond5/6

BenchmarkScorePercentileRankLeader · source

GPQA Diamond

Artificial Analysis v4.3.2 · Oct 27

89.5%

#53 / 147

Leader gpt-6-astra · 96.1%
Artificial Analysis v4.3.2 · Oct 27

SciCode

Artificial Analysis v4.3.2 · Oct 27

49.7%

#68 / 167

Leader opus-5.5 · 66.9%
Artificial Analysis v4.3.2 · Oct 27

CritPt

Artificial Analysis v4.3.2 · Oct 27

8.3%

#71 / 159

Leader Muse Wonder Full · 32.9%
Artificial Analysis v4.3.2 · Oct 27

Humanity's Last Exam

Artificial Analysis v4.3.2 · Oct 27

33.3%

#78 / 167

Leader opus-5.5 · 61.4%
Artificial Analysis v4.3.2 · Oct 27

MMMU-Pro

Artificial Analysis v4.3.2 · Oct 27

74.0%

#60 / 89

Leader fable-5.1 · 90.6%
Artificial Analysis v4.3.2 · Oct 27

Not yet reported

MMLU-Pro

Programmatic Reasoning IQ 107 · measuredBest: #11 of 61 on LiveCodeBench4/6

BenchmarkScorePercentileRankLeader · source

LiveCodeBench

Vals.ai · Aug 4

85.9%

#11 / 61

Leader Fugu Ultra v1.0 · 93.2%
Vals.ai · Aug 4

Terminal-Bench 4.0

Artificial Analysis v4.3.2 · Oct 27

1.0%

#59 / 104

Leader Kimi K3.1 Pro · 67.2%
Artificial Analysis v4.3.2 · Oct 27

ProgramBench

Full source leaderboard reconciliation · Oct 23

0.5%

#48 / 61

Leader opus-5.5 · 65.0%
Full source leaderboard reconciliation · Oct 23

IOI V2

Full source leaderboard reconciliation v2 · Oct 27

9.3%

#42 / 43

Leader Gemini 4 Argon · 100%
Full source leaderboard reconciliation v2 · Oct 27

Not yet reported

SWE-rebenchMirrorCode

Computer Use IQ 114 · measuredBest: #12 of 32 on MCP Atlas3/5

BenchmarkScorePercentileRankLeader · source

MCP Atlas

Full source leaderboard reconciliation · Oct 23

79.2%

#12 / 32

Leader muse-spark-1.1 · 88.1%
Full source leaderboard reconciliation · Oct 23

BrowseComp

Full source leaderboard reconciliation · Oct 23

77.4%

#22 / 43

Leader Kimi K3.1 Pro · 92.1%
Full source leaderboard reconciliation · Oct 23

Arena.ai Agent Arena

Arena.ai Agent Arena · Oct 26

-17.0

#69 / 70

Leader opus-5.5 · 20.9
Arena.ai Agent Arena · Oct 26

Not yet reported

OSWorld-VerifiedAgents' Last Exam

Reliability IQ 106 · measuredBest: #86 of 158 on AA Omniscience3/6

BenchmarkScorePercentileRankLeader · source

AA Omniscience

Artificial Analysis v4.3.2 · Oct 27

-8.9

#86 / 158

Leader opus-5.5 · 46.4
Artificial Analysis v4.3.2 · Oct 27

AA Long Context Reasoning v1.1

Artificial Analysis v4.3.2 · Oct 27

75.7%

#87 / 154

Leader kimi-k3 · 88.7%
Artificial Analysis v4.3.2 · Oct 27

SimpleQA Verified

Full source leaderboard reconciliation · Oct 23

19.1%

#74 / 80

Leader Gemini 4 Argon · 76.0%
Full source leaderboard reconciliation · Oct 23

Not yet reported

BullshitBench v2IFBenchFACTS Grounding

Knowledge Work IQ 114 · measuredBest: #57 of 124 on GDPval-AA v2.1 normalized score5/6

BenchmarkScorePercentileRankLeader · source

GDPval-AA v2.1 normalized score

Artificial Analysis v4.3.2 · Oct 23

31.1%

#57 / 124

Leader opus-5.5 · 68.3%
Artificial Analysis v4.3.2 · Oct 23

GDP.pdf

Artificial Analysis v4.3.2 · Oct 27

11.2%

#56 / 106

Leader Muse Wonder Full · 32.6%
Artificial Analysis v4.3.2 · Oct 27

AA-Briefcase v1.1 Elo

Artificial Analysis v4.3.2 · Oct 27

912

#60 / 111

Leader sonnet-5.5 · 1,812
Artificial Analysis v4.3.2 · Oct 27

EnterpriseOps-Gym-AA

Full source leaderboard reconciliation v1117-task-strict-2026-09-29 · Oct 27

42.7%

#19 / 31

Leader fable-5 · 51.1%
Full source leaderboard reconciliation v1117-task-strict-2026-09-29 · Oct 27

AutomationBench-AA

Artificial Analysis v4.3.2 · Oct 27

4.8%

#78 / 106

Leader Gemini 4 Argon · 77.5%
Artificial Analysis v4.3.2 · Oct 27

Not yet reported

AA-AnalystAgent

Release & availability

Facts from the publisher's announcement and documentation.

Released: Jul 30, 2026
Developer: Thinking Machines
Access: Full weights on Hugging Face; fine-tuning and multimodal chat on Tinker; beta serverless inference and third-party API access
Weights: Open weightsDownloadable and self-hostable
Inputs: textimageaudio
Context: 1,000,000 tokens
API price: $0.30 in · $1.20 outper 1M tokens · full pricing

What Thinking Machines says

The sparse mixture-of-experts model has 276B total parameters and 12B active parameters.
It supports variable thinking effort and a context window of up to 1M tokens.
Thinking Machines reports comparable performance to Inkling at roughly one quarter of the total and active parameter count.

Limitations

The model may hallucinate, miss instructions, and degrade during long multi-turn conversations.
The model card reports occasional compliance with role-play or indirectly framed harmful prompts and recommends defense-in-depth safeguards.

Publisher statements, not AI IQ measurements.

Sourcesthinkingmachines.ai/news/inkling-small thinkingmachines.ai/model-card/inkling-smallhuggingface.co/thinkingmachines/Inkling-Smalltinker-docs.thinkingmachines.ai/tinker/models

About this score

How much of Inkling-Small's IQ is measured, and how the rest is estimated.

27/ 39benchmark results

7/ 7dimensions measured

0estimated

4/5Abstract

3/5Math

5/6Academic

4/6Coding

3/5Computer Use

3/6Reliability

5/6Knowledge Work

Bar height = share of benchmarks with resultsEstimatedNo score

Measured results are shown as published

Every score in the benchmark tables comes from its source unchanged. Composite IQ averages all seven dimensions equally.

How gaps are handled

Every dimension has at least one published result.

A missing result is not a zero

Benchmark versions and evaluation harnesses can differ between sources, so small gaps between models may not be meaningful.

Read the scoring methodology →

Compare Inkling-Small

Side-by-side pages for the models it is most often weighed against.

#1 overall

gpt-6.1-sol

IQ141 31 higherEffective cost$6.30 5.1× pricier

IQ110 same IQEffective cost$0.82 1.5× cheaper

Most similar profile

kimi-k2.7-code

IQ111 1 higherEffective cost$3.57 2.9× pricier

Similar price

kat-coder-pro-v1

IQ103 7 lowerEffective cost$1.27 about the same

#2 overall

opus-5.5

IQ141 31 higherEffective cost$35.73 28.9× pricier

Inkling-SmallComparison model

Compare with any modelBrowse all comparisons →

Frequently asked questions

Quick answers about Inkling-Small, drawn from the same data as the rest of this page.

How AI IQ scores models →Get this data from the API →

What is the IQ of Inkling-Small?: Inkling-Small has an estimated IQ of 110, which ranks #81 out of 176 public models on the AI IQ leaderboard. The score combines seven equally weighted reasoning dimensions using measured benchmark results and conservative estimates for missing coverage.
What is Inkling-Small best at?: Among the dimensions with published benchmark results, Inkling-Small ranks highest in Academic Reasoning (IQ 125, #74 of 203) and Knowledge Work (IQ 114, #65 of 176). Its lowest-ranked measured dimension is Reliability (IQ 106, #124 of 177).
What is the EQ of Inkling-Small?: Inkling-Small doesn't have an EQ score yet. EQ (emotional reasoning) is calculated from EQ-Bench 4 and AttuneBench, and neither has published a result for Inkling-Small. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.
Which benchmarks has Inkling-Small been tested on?: Inkling-Small has published results on 27 of the 39 benchmarks behind AI IQ's Composite IQ, including LiveCodeBench, GPQA Diamond, MCP Atlas, SciCode and CritPt.
How much does Inkling-Small cost?: Inkling-Small costs $0.30 per million input tokens and $1.20 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $1.24 per million input-plus-output tokens.
Is Inkling-Small open source?: Yes — Inkling-Small is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.
Who developed Inkling-Small?: Inkling-Small was developed by Thinking Machines, based in the United States.
When was Inkling-Small released?: Inkling-Small was publicly released on July 30, 2026.
How large is the context window of Inkling-Small?: Inkling-Small supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.
How fast is Inkling-Small?: In independent measurements, Inkling-Small generates about 224 tokens per second, with a typical response time of about 13 seconds.