What is the IQ of Inkling 2 Small?

Inkling 2 Small has an estimated IQ of 126, which ranks #27 out of 180 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 2 Small best at?

Among the dimensions with published benchmark results, Inkling 2 Small ranks highest in Abstract Reasoning (IQ 123, #14 of 209) and Computer Use (IQ 128, #20 of 180). Its lowest-ranked measured dimension is Academic Reasoning (IQ 128, #60 of 211).

What is the EQ of Inkling 2 Small?

Inkling 2 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 2 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 2 Small been tested on?

Inkling 2 Small has published results on 12 of the 39 benchmarks behind AI IQ's Composite IQ, including GDP.pdf, MCP Atlas, BrowseComp, AA-Briefcase v1.1 Elo and ARC-AGI-1.

How much does Inkling 2 Small cost?

Inkling 2 Small costs $0.30 per million input tokens and $1.00 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $1.25 per million input-plus-output tokens.

Is Inkling 2 Small open source?

No — Inkling 2 Small is a proprietary model. Its weights are not publicly available, so it is used through hosted APIs and products.

Who developed Inkling 2 Small?

Inkling 2 Small was developed by Thinking Machines, based in the United States.

When was Inkling 2 Small released?

Inkling 2 Small was publicly released on November 5, 2026.

How large is the context window of Inkling 2 Small?

AI IQ hasn't verified an input context limit for Inkling 2 Small from its developer's documentation, so none is listed rather than guessing.

How fast is Inkling 2 Small?

AI IQ doesn't have independent speed measurements for Inkling 2 Small yet. Speed comes from independent tests of a model's hosted API, which aren't available for every model — especially new, restricted-access, or self-hosted ones.

Thinking MachinesReleased Nov 5, 2026

NewProprietary

Inkling 2 Small

Strongest in Abstract Reasoning (#14 of 209) and Computer Use (#20 of 180); weakest in Academic Reasoning (#60 of 211). Reliability is estimated, because no results are published yet.

Composite IQ ?

126

#27 of 180 ranked models · ahead of 85%

US (Other)'s top-ranked model

Effective cost

$1.25/1M

$0.30 in · $1 out

Speed

Not measured

No speed data yet

Context

Unverified

Not in publisher docs

Evidence

6of 7 dims

12 of 39 benchmarks

Compare with fable-5.2API pricing

IQ profile

Seven dimensions, averaged equally into Composite IQ

Compare shape with

MeasuredEstimated95% 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

123measured

#14 / 209

2/5

-9

Mathematical ReasoningMath

128measured

#37 / 180

1/5

-19

Academic ReasoningAcademic

128measured

#60 / 211

1/6

-7

Programmatic ReasoningCoding

135measured

#34 / 181

2/6

-15

Computer UseComputer Use

128measured

#20 / 180

3/5

-16.5

ReliabilityReliability

112estimated

#96 / 181

0/6

-22.5

Knowledge WorkKnowledge Work

128measured

#23 / 180

3/6

-4

Inkling 2 Small95% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human

Benchmark results

12 of 39 Composite IQ benchmarks have a published result for Inkling 2 Small. Benchmarks without a result stay listed so gaps are visible.

Abstract Reasoning IQ 123 · measuredBest: #14 of 71 on ARC-AGI-12/5

BenchmarkScorePercentileRankLeader · source

ARC-AGI-1

Full source leaderboard reconciliation · Nov 13

96.5%

#14 / 71

Leader fable-5 · 98.5%
Full source leaderboard reconciliation · Nov 13

ARC-AGI-2

Full source leaderboard reconciliation · Nov 13

81.7%

#17 / 71

Leader gpt-6-astra · 95.0%
Full source leaderboard reconciliation · Nov 13

Not yet reported

ARC-AGI-3Chess PuzzlesMystery Game Puzzles

Mathematical Reasoning IQ 128 · measuredBest: #17 of 66 on MathArena1/5

BenchmarkScorePercentileRankLeader · source

MathArena

Full source leaderboard reconciliation · Nov 13

42.3%

#17 / 66

Leader gpt-6.1-sol · 80.6%
Full source leaderboard reconciliation · Nov 13

Not yet reported

FrontierMath Tier 4FrontierMath Tier 1-3ProofBench 1.1AIME

Academic Reasoning IQ 128 · measuredBest: #19 of 90 on MMMU-Pro1/6

BenchmarkScorePercentileRankLeader · source

MMMU-Pro

Artificial Analysis v5.0.1 · Nov 13

81.6%

#19 / 90

Leader fable-5.1 · 90.6%
Artificial Analysis v5.0.1 · Nov 13

Not yet reported

Humanity's Last ExamGPQA DiamondCritPtSciCodeMMLU-Pro

Programmatic Reasoning IQ 135 · measuredBest: #40 of 109 on Terminal-Bench 4.02/6

BenchmarkScorePercentileRankLeader · source

Terminal-Bench 4.0

Artificial Analysis v5.0.1 · Nov 13

13.1%

#40 / 109

Leader Kimi K3.1 Pro · 67.2%
Artificial Analysis v5.0.1 · Nov 13

ProgramBench

Full source leaderboard reconciliation · Nov 6

12.0%

#27 / 63

Leader opus-5.5 · 65.0%
Full source leaderboard reconciliation · Nov 6

Not yet reported

LiveCodeBenchIOI V2SWE-rebenchMirrorCode

Computer Use IQ 128 · measuredBest: #5 of 35 on MCP Atlas3/5

BenchmarkScorePercentileRankLeader · source

MCP Atlas

Scale Labs MCP Atlas · Nov 10

85.4%

#5 / 35

Leader Gemini 4 Argon · 90.7%
Scale Labs MCP Atlas · Nov 10

BrowseComp

Full source leaderboard reconciliation · Nov 9

90.2%

#7 / 45

Leader Kimi K3.1 Pro · 92.1%
Full source leaderboard reconciliation · Nov 9

Arena.ai Agent Arena

Arena.ai Agent Arena · Nov 13

2.5

#21 / 72

Leader fable-5.2 · 30.4
Arena.ai Agent Arena · Nov 13

Not yet reported

OSWorld-VerifiedAgents' Last Exam

Reliability IQ 112 · estimatedNo published results0/6

No published results yet for Reliability. The 112 shown is estimated from Inkling 2 Small's other dimensions and will be replaced as results come in. How estimates work

SimpleQA VerifiedaaOmniscienceBullshitBench v2IFBenchAA Long Context Reasoning v1.1FACTS Grounding

Knowledge Work IQ 128 · measuredBest: #15 of 111 on GDP.pdf3/6

BenchmarkScorePercentileRankLeader · source

GDP.pdf

Artificial Analysis v5.0.1 · Nov 13

25.2%

#15 / 111

Leader Muse Wonder Full · 32.6%
Artificial Analysis v5.0.1 · Nov 13

AA-Briefcase v1.1 Elo

Artificial Analysis v5.0.1 · Nov 13

1,527

#21 / 116

Leader sonnet-5.5 · 1,816
Artificial Analysis v5.0.1 · Nov 13

AutomationBench-AA

Artificial Analysis v5.0.1 · Nov 13

45.6%

#48 / 112

Leader Gemini 4 Carbon · 77.8%
Artificial Analysis v5.0.1 · Nov 13

Not yet reported

GDPval-AA v2.1AA-AnalystAgentEnterpriseOps-Gym-AA

Release & availability

Facts from the publisher's announcement and documentation.

Released: Nov 5, 2026
Developer: Thinking Machines
Access: Available now for sampling, fine-tuning and playground use through Tinker as thinkingmachines/Inkling-2-Small. Weights are closed at launch.
Weights: ProprietaryHosted access only
Inputs: textimage
Context: Not publishedThe model card says up to 1 million tokens; the current Tinker catalog lists 64K and 256K deployment variants. The available input limit depends on the selected Tinker variant.
API price: $0.30 in · $1 outper 1M tokens · full pricing

What Thinking Machines says

Second-generation mixture-of-experts model with 594 billion total parameters and 17 billion active parameters.
Accepts text and image inputs, generates text, and supports variable reasoning effort.
Current Tinker sampling prices are $0.30 per million prefill tokens and $1.00 per million output tokens for the listed variants.

Limitations

The launch release keeps weights closed; an Apache 2.0 weight release is planned after external safety testing, but has not occurred.
The 1 million-token model-card capability exceeds the currently listed 64K and 256K Tinker variants; no single 1 million-token hosted endpoint is verified.
Publisher evaluation scores use their stated model settings and harnesses; they are not imported as independent AI IQ benchmark observations without exact source mapping.

Publisher statements, not AI IQ measurements.

Sourcesthinkingmachines.ai/news/inkling-2-small thinkingmachines.ai/model-card/inkling-2-smalltinker-docs.thinkingmachines.ai/tinker/models/models_and_pricing

About this score

How much of Inkling 2 Small's IQ is measured, and how the rest is estimated.

12/ 39benchmark results

6/ 7dimensions measured

1estimated

2/5Abstract

1/5Math

1/6Academic

2/6Coding

3/5Computer Use

0/6Reliability

3/6Knowledge 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.

Estimates fill the gaps

Reliability has no published results yet, so its IQ is estimated from Inkling 2 Small's other dimensions.

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 2 Small

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

#1 overall

fable-5.2

IQ144 18 higherEffective cost$73.69 58.7× pricier

Closest IQ

gemini-3.7-flash

IQ126 same IQEffective cost$5.96 4.8× pricier

Most similar profile

DeepSeek V4.1 Flash

IQ125 1 lowerEffective cost$1.85 1.5× pricier

Similar price

GLM-5.3-Flash

IQ119 7 lowerEffective cost$1.31 about the same

IQ141 15 higherEffective cost$6.34 5.1× pricier

Inkling 2 SmallComparison model

Compare with any modelBrowse all comparisons →

Frequently asked questions

Quick answers about Inkling 2 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 2 Small?: Inkling 2 Small has an estimated IQ of 126, which ranks #27 out of 180 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 2 Small best at?: Among the dimensions with published benchmark results, Inkling 2 Small ranks highest in Abstract Reasoning (IQ 123, #14 of 209) and Computer Use (IQ 128, #20 of 180). Its lowest-ranked measured dimension is Academic Reasoning (IQ 128, #60 of 211).
What is the EQ of Inkling 2 Small?: Inkling 2 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 2 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 2 Small been tested on?: Inkling 2 Small has published results on 12 of the 39 benchmarks behind AI IQ's Composite IQ, including GDP.pdf, MCP Atlas, BrowseComp, AA-Briefcase v1.1 Elo and ARC-AGI-1.
How much does Inkling 2 Small cost?: Inkling 2 Small costs $0.30 per million input tokens and $1.00 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $1.25 per million input-plus-output tokens.
Is Inkling 2 Small open source?: No — Inkling 2 Small is a proprietary model. Its weights are not publicly available, so it is used through hosted APIs and products.
Who developed Inkling 2 Small?: Inkling 2 Small was developed by Thinking Machines, based in the United States.
When was Inkling 2 Small released?: Inkling 2 Small was publicly released on November 5, 2026.
How large is the context window of Inkling 2 Small?: AI IQ hasn't verified an input context limit for Inkling 2 Small from its developer's documentation, so none is listed rather than guessing.
How fast is Inkling 2 Small?: AI IQ doesn't have independent speed measurements for Inkling 2 Small yet. Speed comes from independent tests of a model's hosted API, which aren't available for every model — especially new, restricted-access, or self-hosted ones.