Skip to main content
BenchLM
Data

Inkling vs Inkling-Small

Updated October 29, 2026. Rank says Inkling-Small is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty. This is a same-family comparison, so migration details appear when the source data supports them.

Share or export
Share on XLinkedInAPI/MCP

Decision reading

Inkling-Small has the higher public score estimate, 49.5 versus 47.91, but the 90% score intervals overlap. Treat that as a lead, not a settled winner. 22 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.

Model A
Thinking Machines Lab logo
Inkling

Thinking Machines Lab

47.91/100

Supported · Public rank #108

90% interval 29.8–66.0

Model B
Thinking Machines Lab logo
Inkling-Small

Thinking Machines Lab

49.5/100

Supported · Public rank #96

90% interval 32.4–66.6

Shared results
22
Inkling only
2
Inkling-Small only
6
Like-for-like categories
5 / 8
Supported: Inkling and Inkling-SmallHow the comparison works

Which one for your work

Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.

  • Coding work

    Code generation, repair, and software-engineering tasks

    Inkling-Small

    Inkling-Small has the higher public coding point estimate, 36.9 to 33, with Supported evidence for both models. The conditional ranges do not establish rank confidence.

    Confidence: limited

  • Agentic work

    Tool use, computer use, and multi-step task completion

    Inkling

    Inkling has the higher public agentic point estimate, 37 to 36.1, with Supported evidence for both models. The conditional ranges do not establish rank confidence.

    Confidence: limited

  • Chat turn cost

    1K fresh input + 500 output tokens

    Inkling-Small

    Inkling-Small has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.

    Confidence: listed-rates
Show secondary and unsupported calls

  • Cache-heavy agent loop cost

    200K cached + 20K fresh input + 10K output tokens

    Inkling-Small

    Inkling-Small has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.

    Confidence: listed-rates

  • Repository review cost

    50K fresh input + 3K output tokens

    Inkling-Small

    Inkling-Small has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.

    Confidence: listed-rates

  • Long documents

    Prompts that approach the documented context limit

    No clear pick

    The documented context windows are equal.

    Confidence: documented

Which one for a specific job

Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.

The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.

33.0Inkling36.9Inkling-Small

Like-for-like · BenchAlign v5.8

Inkling-Small has the higher coding point estimate. Conditional score ranges do not establish rank confidence.

Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.

Open the coding leaderboardShortlist models for this job
Same basis rules as the category table below

What is actually comparable

Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.

1 category rests on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.

Shape of the matched evidence

Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.

Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.

Bars run 0–100 on each benchmark’s normalized display scale

Category results, on a stated basis

Each row shows the public-lane category score for both models: the BenchAlign v5.8 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.

Agentic

Like-for-like
Inkling
37.0
Supported · #81/135
Inkling-Small
36.1
Supported · #86/135
Basis
BenchAlign v5.8 lane · 5 vs 5 public rows
Reading
Inkling leads · intervals overlap

Coding

Like-for-like
Inkling
33.0
Supported · #90/156
Inkling-Small
36.9
Supported · #77/156
Basis
BenchAlign v5.8 lane · 6 vs 6 public rows
Reading
Inkling-Small leads · intervals overlap

Multimodal

Like-for-like
Inkling
68.0
#38/58
Inkling-Small
67.3
#41/58
Basis
Provisional lane · 2 vs 2 weighted rows
Reading
Inkling leads

Knowledge

Like-for-like
Inkling
50.6
Supported · #76/179
Inkling-Small
50.6
Supported · #75/179
Basis
BenchAlign v5.8 lane · 6 vs 6 public rows
Reading
Tie

Instruction following

Like-for-like
Inkling
87.6
#40/157
Inkling-Small
93.3
#10/157
Basis
Provisional lane · 1 vs 1 weighted rows
Reading
Inkling-Small leads

Math

Directional only
Inkling
58.3
#11/12
Inkling-Small
69.8
#5/12
Basis
Provisional lane · 3 vs 4 weighted rows
Reading
Directional only

Reasoning

Not comparable
Inkling
78.1
#11/26
Inkling-Small
55.0
Unranked · 4 rankable rows
Basis
Provisional lane · 0 vs 1 weighted rows
Reading
Not comparable

Multilingual

Not comparable
Inkling
Not ranked
Inkling-Small
Not ranked
Basis
Provisional lane · 0 vs 0 weighted rows
Reading
Not comparable

Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.8) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.

Supported evidence per lane · bars run 0–100Methodology

What each workload costs

Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.

Chat turn

1K fresh input + 500 output tokens

Inkling
$0.00421
Fits in one request
Inkling-Small
$0.0013
Fits in one request

Inkling-Small has the lower modeled cost

Costs use the listed standard API rates.

Repository review

50K fresh input + 3K output tokens

Inkling
$0.10754
Fits in one request
Inkling-Small
$0.03332
Fits in one request

Inkling-Small has the lower modeled cost

Costs use the listed standard API rates.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

Inkling
$0.159
Fits in one request
Inkling-Small
$0.0492
Fits in one request

Inkling-Small has the lower modeled cost

Costs use the listed standard API rates.

Cached input falls back to the list input rate only where a cached rate is unpublished

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

Context window

Maximum documented context; output-token limits may be lower.

Inkling

1M

Inkling-Small

1M

API model ID

Inkling

Not sourced

Inkling-Small

Not sourced

Cached-input rate

A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.

Inkling

$0.374 per 1M cached input tokens

Inkling-Small

$0.116 per 1M cached input tokens

Documented inputs

Inkling

Not sourced

Inkling-Small

Not sourced

Documented outputs

Inkling

Not sourced

Inkling-Small

Not sourced

Provider availability

Inkling

Not sourced

Inkling-Small

Not sourced

Reasoning profile

Inkling

Hybrid

Inkling-Small

Hybrid

Weight access

Inkling

Open Weight

Inkling-Small

Open Weight

License

Inkling

Open Weight

Inkling-Small

Open Weight

Release date

Inkling

2026-07-15

Inkling-Small

2026-07-30

If you are choosing between sibling variants

Deployment change
Both entries list Thinking Machines Lab as the provider. Confirm endpoint, model ID, limits, and feature support before switching.
Quality signal
Inkling-Small has the higher public score estimate, 49.5 versus 47.91, but the 90% score intervals overlap.
Workload cost
Repository review: $0.10754 vs $0.03332. Cache-heavy agent loop: $0.159 vs $0.0492.
Context tradeoff
Both models list 1M.
Run the same representative tasks against both endpoints before changing production traffic.

Questions

Which is better, Inkling or Inkling-Small?

Inkling-Small has the higher public score estimate, 49.5 versus 47.91, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.

Which is better for coding, Inkling or Inkling-Small?

Inkling-Small has the higher public coding point estimate, 36.9 to 33, with Supported evidence for both models. The conditional ranges do not establish rank confidence.

Which is better for agentic tasks, Inkling or Inkling-Small?

Inkling has the higher public agentic tasks point estimate, 37 to 36.1, with Supported evidence for both models. The conditional ranges do not establish rank confidence.

Which costs less, Inkling or Inkling-Small?

For the stated presets, chat costs $0.00421 on Inkling and $0.0013 on Inkling-Small; repository review costs $0.10754 and $0.03332; the cache-heavy agent loop costs $0.159 and $0.0492. Costs use the listed standard API rates.

Which has the larger context window, Inkling or Inkling-Small?

Both models list the same context window, 1M.

Self-host vs API cost

Estimates at 50,000 req/day · 1000 tokens/req average.

Inkling
API / mo$4,913
Self-host / mo$27,962
Break-even344M/day
Inkling-Small
API / mo$1,515
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Model the full break-even

Benchmark evidence

The full public result ledger is available for audit without forcing a wide desktop table onto a phone.

Browse raw public benchmark evidence30 rows

Agentic

  • Terminal-Bench 2.1

    Inkling63.8%
    Source
    Inkling-Small64.7%
    Source

    Inkling-Small leads this result

  • BrowseComp

    Inkling77.1%
    Source
    Inkling-Small77.4%
    Source

    Inkling-Small leads this result

  • MCP Atlas

    Inkling74.1%
    Source
    Inkling-Small79.6%
    Source

    Inkling-Small leads this result

  • Terminal-Bench 2.1 (Vals)

    Inkling47.6%
    Source
    Inkling-Small55.1%
    Source

    Inkling-Small leads this result

  • CWE-bench v1

    Inkling37.0%
    Source
    Inkling-Small—

    Not directly comparable

  • Toolathlon-Verified

    Inkling—
    Inkling-Small54.4%
    Source

    Not directly comparable

Coding

  • SWE-bench Verified

    Inkling77.6%
    Source
    Inkling-Small80.2%
    Source

    Inkling-Small leads this result

  • SWE-bench Pro

    Inkling54.3%
    Source
    Inkling-Small55.9%
    Source

    Inkling-Small leads this result

  • Terminal-Bench 2.1

    Inkling63.8%
    Source
    Inkling-Small64.7%
    Source

    Inkling-Small leads this result

  • FrontierSWE v2

    Inkling4.1%
    Source
    Inkling-Small—

    Not directly comparable

  • LiveCodeBench (Vals)

    Inkling85.5%
    Source
    Inkling-Small85.9%
    Source

    Inkling-Small leads this result

  • SWE-bench (Vals)

    Inkling77.6%
    Source
    Inkling-Small82.2%
    Source

    Inkling-Small leads this result

  • SciCode

    Inkling—
    Inkling-Small48.7%
    Source

    Not directly comparable

Reasoning

  • ARC-AGI-2

    Inkling—
    Inkling-Small40.1%
    Source

    Not directly comparable

  • CritPt

    Inkling—
    Inkling-Small8.3%
    Source

    Not directly comparable

  • ARC-AGI-1

    Inkling—
    Inkling-Small84.00%
    Source

    Not directly comparable

Multimodal

  • MMMU-Pro

    Inkling73.5%
    Source
    Inkling-Small74%
    Source

    Inkling-Small leads this result

  • CharXiv

    Inkling82%
    Source
    Inkling-Small81.3%
    Source

    Inkling leads this result

  • CharXiv w/o tools

    Inkling78.1%
    Source
    Inkling-Small77.4%
    Source

    Inkling leads this result

Knowledge

  • GPQA

    Inkling87.9%
    Source
    Inkling-Small89.5%
    Source

    Inkling-Small leads this result

  • GPQA-D

    Inkling87.9%
    Source
    Inkling-Small89.5%
    Source

    Inkling-Small leads this result

  • HLE

    Inkling46%
    Source
    Inkling-Small47.8%
    Source

    Inkling-Small leads this result

  • HLE w/o tools

    Inkling30%
    Source
    Inkling-Small31.6%
    Source

    Inkling-Small leads this result

  • GPQA Diamond (Vals)

    Inkling87.1%
    Source
    Inkling-Small83.6%
    Source

    Inkling leads this result

  • MMLU-Pro (Vals)

    Inkling86.3%
    Source
    Inkling-Small85.6%
    Source

    Inkling leads this result

Instruction following

  • IFBench

    Inkling79.8%
    Source
    Inkling-Small82.2%
    Source

    Inkling-Small leads this result

Math

  • AIME26

    Inkling97.1%
    Source
    Inkling-Small95.5%
    Source

    Inkling leads this result

  • FrontierMath v2 (Tiers 1-3)

    Shared source
    Inkling33.333%
    Inkling-Small46.316%

    Inkling-Small leads this result

  • FrontierMath v2 (Tier 4)

    Shared source
    Inkling4.878%
    Inkling-Small17.073%

    Inkling-Small leads this result

  • HMMT Feb 2026

    Inkling—
    Inkling-Small90.2%
    Source

    Not directly comparable

30 public results · 22 shared

Related comparisons

Compare API pricingRead the methodologyOpen the model selector

Watch Inkling vs Inkling-Small

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 6,100+ subscribers.

Last updated October 29, 2026