Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d261Policy30d65, 14 pendingSources1,013

Inkling

Thinking MachinesReleased Jul 15, 2026Current

Score basis

Inkling is capable in long context, instruction following, and factuality; and behind the leaders in reasoning, agentic tasks, multimodal tasks, coding, and math. Too few results yet to rate safety or multilingual tasks.

Capability profile

Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.

Capable3 capabilities
  1. Long Context−16.0%1 of 3
  2. Instruction Following−18.3%2 of 3
  3. Factuality−22.1%3 of 4
Limited5 capabilities
  1. Reasoning−27.5%6 of 6
  2. Agentic−29.3%5 of 7
  3. Multimodal−30.8%2 of 6
  4. Coding−36.9%6 of 10
  5. Math−52.2%3 of 5
Not rated2 capabilities

Too few results yet to rate Safety or Multilingual.

Price

$1.00input$4.05outputper million tokens

From Artificial Analysis · 4 providers tracked · All prices

Costs more than 69% of 331 priced models · 3:1 input-to-output blend, log scale

Evidence

80results on53benchmarks

  • 17 independently verified
  • 16 aggregator
  • 47 vendor-reported

From 16 sources · latest Oct 8, 2026 · How verification works

API features

Tool callingReasoning

As listed by OpenRouter

Research

2 papers reference Inkling

Inkling benchmark results

80 results on 53 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.

Long Context

CapableFull long context ranking

16.0% behind the leader1 of 3 ranked benchmarks measured

Instruction Following

CapableFull instruction following ranking

18.3% behind the leader2 of 3 ranked benchmarks measured

Show 1 more instruction following resultHide 1 instruction following result

Factuality

CapableFull factuality ranking

22.1% behind the leader3 of 4 ranked benchmarks measured

Show 2 more factuality resultsHide 2 factuality results

Reasoning

LimitedFull reasoning ranking

27.5% behind the leader6 of 6 ranked benchmarks measured

Show 5 more reasoning resultsHide 5 reasoning results

Agentic

LimitedFull agentic ranking

29.3% behind the leader5 of 7 ranked benchmarks measured

Show 3 more agentic resultsHide 3 agentic results

Multimodal

LimitedFull multimodal ranking

30.8% behind the leader2 of 6 ranked benchmarks measured

Show 1 more multimodal resultHide 1 multimodal result

Coding

LimitedFull coding ranking

36.9% behind the leader6 of 10 ranked benchmarks measured

Show 3 more coding resultsHide 3 coding results

Math

LimitedFull math ranking

52.2% behind the leader3 of 5 ranked benchmarks measured

Show 3 more math resultsHide 3 math results

More results

Benchmarks outside the capability baskets. They are not ranked against a leader.

Show 29 more resultsHide 29 results

Inkling: common questions

Who makes Inkling?

Inkling is made by Thinking Machines.

When was Inkling released?

Inkling was released on Jul 15, 2026, according to Thinking Machines's own announcement.

What is Inkling good at?

Inkling is capable in long context, instruction following, and factuality; and behind the leaders in reasoning, agentic tasks, multimodal tasks, coding, and math. Too few results yet to rate safety or multilingual tasks.

How much does Inkling cost?

Inkling costs $1.00 per million input tokens and $4.05 per million output tokens, according to Artificial Analysis. We track its price at 4 providers. At a mix of three input tokens to one output token, it costs more than 69% of the 331 priced models we track.

How many benchmarks has Inkling been tested on?

We track 80 results for Inkling on 53 benchmarks from 16 sources, 17 of them independently verified. The latest was recorded on Oct 8, 2026.

Which API features does Inkling support?

OpenRouter lists tool calling and reasoning for Inkling.

About this record

Where Inkling's numbers come from, and every name it appears under.

Tracked since
Jul 15, 2026
Newest source mention
Sep 16, 2026

Where the results come from

Verification: 80 scores · 17 independently verified · 16 aggregator-attributed · 47 vendor-reported. How these tiers are assigned

From 16 sources on 11 sites. thinkingmachines.ai supplies 19 of them; the 17 independently verified results come from 8 sites. Bars are coloured by trust tier.

  • thinkingmachines.ai19
  • artificialanalysis.ai17
  • huggingface.co17
  • api.llm-stats.com11
  • livebench.ai7
  • epoch.ai3
  • arcprize.org2
  • datasets-server.huggingface.co1
  • labs.scale.com1
  • lmarena.ai1
  • simple-bench.com1

Also known as

inkling (xhigh)inkling:batchthinking machines inklingthinkingmachines/inklingthinking machines' inklinginkling:freeinkling xhigh