Skip to main content
BenchLM
Data

Inkling vs Qwen3.5 397B

Updated October 14, 2026. Rank says Qwen3.5 397B is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.

Share or export
Share on XLinkedInAPI/MCP

Decision reading

Qwen3.5 397B has the higher public point estimate, 55.63 versus 52.13. Their conditional score ranges overlap. These ranges do not establish rank confidence. 10 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

52.13/100

Supported · Public rank #85

90% interval 41.7–62.5

Model B
Alibaba logo
Qwen3.5 397B

Alibaba

55.63/100

Estimated · Public rank #63

Conditional range 41.3–70.0

Shared results
10
Inkling only
14
Qwen3.5 397B only
52
Like-for-like categories
2 / 8
Supported: Inkling · Estimated: Qwen3.5 397B. Conditional ranges do not establish rank confidence.How 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.

  • Long documents

    Prompts that approach the documented context limit

    Inkling

    Inkling has the larger documented context window.

    Confidence: documented

  • Chat turn cost

    1K fresh input + 500 output tokens

    Qwen3.5 397B

    Qwen3.5 397B 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

  • Repository review cost

    50K fresh input + 3K output tokens

    Qwen3.5 397B

    Qwen3.5 397B has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.

    Confidence: listed-rates

  • Coding work

    Code generation, repair, and software-engineering tasks

    Not enough matched evidence

    Qwen3.5 397B is not ranked on the public lane for coding, so no winner is named for coding.

    Confidence: limited

  • Agentic work

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

    Not enough matched evidence

    Qwen3.5 397B is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.

    Confidence: limited

  • Cache-heavy agent loop cost

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

    Not enough matched evidence

    The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. Qwen3.5 397B does not fit this workload in one request. Qwen3.5 397B has no published cached-input rate, so cached tokens use its listed input rate.

    Confidence: rate-fallback

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.

34.0Inkling—Qwen3.5 397B

Not comparable · BenchAlign v5.8

The coding row is not comparable on the public lane: at least one model is not measured or not ranked there.

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.

2 categories rest 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.

Multimodal

Like-for-like
Inkling
67.5
#35/54
Qwen3.5 397B
73.6
#29/54
Basis
Provisional lane · 2 vs 2 weighted rows
Reading
Qwen3.5 397B leads

Instruction following

Like-for-like
Inkling
87.6
#35/128
Qwen3.5 397B
79.9
#56/128
Basis
Provisional lane · 1 vs 1 weighted rows
Reading
Inkling leads

Agentic

Directional only
Inkling
36.2
Supported · #72/124
Qwen3.5 397B
44.3
Estimated · #55/124
Basis
BenchAlign v5.8 lane · 5 vs 14 public rows
Reading
Directional only

Knowledge

Directional only
Inkling
54.5
Supported · #62/181
Qwen3.5 397B
54.1
Estimated · #64/181
Basis
BenchAlign v5.8 lane · 6 vs 6 public rows
Reading
Directional only

Coding

Not comparable
Inkling
34.0
Supported · #80/149
Qwen3.5 397B
Not ranked
Basis
BenchAlign v5.8 lane · 6 vs 6 public rows
Reading
Not comparable

Reasoning

Not comparable
Inkling
79.0
#14/29
Qwen3.5 397B
74.9
Unranked · 2 rankable rows
Basis
Provisional lane · 0 vs 1 weighted rows
Reading
Not comparable

Multilingual

Not comparable
Inkling
Not ranked
Qwen3.5 397B
93.8
#5/17
Basis
Provisional lane · 0 vs 1 weighted rows
Reading
Not comparable

Math

Not comparable
Inkling
58.3
#11/12
Qwen3.5 397B
81.7
Unranked · 5 rankable rows
Basis
Provisional lane · 3 vs 2 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
Qwen3.5 397B
$0.0024
Fits in one request

Qwen3.5 397B 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
Qwen3.5 397B
$0.0408
Fits in one request

Qwen3.5 397B 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
Qwen3.5 397B
$0.168
Does not fit in one request
Cached input priced at the published list-input rate

Qwen3.5 397B does not fit this workload in one request. Qwen3.5 397B has no published cached-input rate, so cached tokens use its listed input rate.

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

Qwen3.5 397B

128K

API model ID

Inkling

Not sourced

Qwen3.5 397B

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

Qwen3.5 397B

Not published

Documented inputs

Inkling

Not sourced

Qwen3.5 397B

Not sourced

Documented outputs

Inkling

Not sourced

Qwen3.5 397B

Not sourced

Provider availability

Inkling

Not sourced

Qwen3.5 397B

Not sourced

Reasoning profile

Inkling

Hybrid

Qwen3.5 397B

Non-Reasoning

Weight access

Inkling

Open Weight

Qwen3.5 397B

Open Weight

License

Inkling

Open Weight

Qwen3.5 397B

Open Weight

Release date

Inkling

2026-07-15

Qwen3.5 397B

2026-02-16

If you already use one of these models

Deployment change
The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
Quality signal
Qwen3.5 397B has the higher public point estimate, 55.63 versus 52.13. Their conditional score ranges overlap. These ranges do not establish rank confidence.
Workload cost
Repository review: $0.10754 vs $0.0408. Cache-heavy agent loop: $0.159 vs $0.168.
Context tradeoff
Inkling has the larger documented window (1M).
Run the same representative tasks against both endpoints before changing production traffic.

Questions

Which is better, Inkling or Qwen3.5 397B?

Qwen3.5 397B has the higher public point estimate, 55.63 versus 52.13. Their conditional score ranges overlap. These ranges do not establish rank confidence. The page therefore keeps the decision tied to the specific documented workload.

Which is better for coding, Inkling or Qwen3.5 397B?

Qwen3.5 397B is not ranked on the public lane for coding, so no winner is named for coding.

Which is better for agentic tasks, Inkling or Qwen3.5 397B?

Qwen3.5 397B scores higher for agentic tasks on the public lane, 44.3 to 36.2. Qwen3.5 397B is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.

Which costs less, Inkling or Qwen3.5 397B?

For the stated presets, chat costs $0.00421 on Inkling and $0.0024 on Qwen3.5 397B; repository review costs $0.10754 and $0.0408; the cache-heavy agent loop costs $0.159 and $0.168. Qwen3.5 397B does not fit this workload in one request. Qwen3.5 397B has no published cached-input rate, so cached tokens use its listed input rate.

Which has the larger context window, Inkling or Qwen3.5 397B?

Inkling has the larger documented context window: 1M, compared with 128K.

Benchmark evidence

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

Browse raw public benchmark evidence76 rows

Agentic

  • Terminal-Bench 2.1

    Inkling63.8%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • BrowseComp

    Inkling77.1%
    Source
    Qwen3.5 397B69%
    Source

    Inkling leads this result

  • MCP Atlas

    Inkling74.1%
    Source
    Qwen3.5 397B46.1%
    Source

    Inkling leads this result

  • Terminal-Bench 2.1 (Vals)

    Inkling47.6%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • CWE-bench v1

    Inkling37.0%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • Terminal-Bench 2.0

    Inkling—
    Qwen3.5 397B52.5%
    Source

    Not directly comparable

  • OSWorld-Verified

    Inkling—
    Qwen3.5 397B62.2%
    Source

    Not directly comparable

  • Claw-Eval

    Inkling—
    Qwen3.5 397B56.8%
    Source

    Not directly comparable

  • QwenClawBench

    Inkling—
    Qwen3.5 397B51.8%
    Source

    Not directly comparable

  • τ³-bench results

    Inkling—
    Qwen3.5 397B68.4%
    Source

    Not directly comparable

  • VITA-Bench

    Inkling—
    Qwen3.5 397B43.7%
    Source

    Not directly comparable

  • DeepPlanning

    Inkling—
    Qwen3.5 397B37.6%
    Source

    Not directly comparable

  • Toolathlon

    Inkling—
    Qwen3.5 397B36.3%
    Source

    Not directly comparable

  • MCP-Tasks

    Inkling—
    Qwen3.5 397B74.2%
    Source

    Not directly comparable

  • WideResearch

    Inkling—
    Qwen3.5 397B74.0%
    Source

    Not directly comparable

  • Gert Labs

    Inkling—
    Qwen3.5 397B46.76%
    Source

    Not directly comparable

  • ResearchClawBench

    Inkling—
    Qwen3.5 397B14.2%
    Source

    Not directly comparable

Coding

  • SWE-bench Verified

    Inkling77.6%
    Source
    Qwen3.5 397B76.2%
    Source

    Inkling leads this result

  • SWE-bench Pro

    Inkling54.3%
    Source
    Qwen3.5 397B50.9%
    Source

    Inkling leads this result

  • Terminal-Bench 2.1

    Inkling63.8%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • FrontierSWE v2

    Inkling4.1%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • LiveCodeBench (Vals)

    Inkling85.5%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • SWE-bench (Vals)

    Inkling77.6%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • LiveCodeBench v6

    Inkling—
    Qwen3.5 397B83.6%
    Source

    Not directly comparable

  • SWE Multilingual

    Inkling—
    Qwen3.5 397B69.3%
    Source

    Not directly comparable

  • NL2Repo

    Inkling—
    Qwen3.5 397B32.2%
    Source

    Not directly comparable

  • Terminal-Bench 2.0

    Inkling—
    Qwen3.5 397B52.5%
    Source

    Not directly comparable

Reasoning

  • LongBench v2

    Inkling—
    Qwen3.5 397B63.2%
    Source

    Not directly comparable

  • AI-Needle

    Inkling—
    Qwen3.5 397B68.7%
    Source

    Not directly comparable

Multimodal

  • MMMU-Pro

    Inkling73.5%
    Source
    Qwen3.5 397B79%
    Source

    Qwen3.5 397B leads this result

  • CharXiv

    Inkling82%
    Source
    Qwen3.5 397B80.8%
    Source

    Inkling leads this result

  • CharXiv w/o tools

    Inkling78.1%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • MMMU

    Inkling—
    Qwen3.5 397B85.0%
    Source

    Not directly comparable

  • RealWorldQA

    Inkling—
    Qwen3.5 397B83.9%
    Source

    Not directly comparable

  • OmniDocBench 1.5

    Inkling—
    Qwen3.5 397B90.8%
    Source

    Not directly comparable

  • Video-MME (with subtitle)

    Inkling—
    Qwen3.5 397B87.5%
    Source

    Not directly comparable

  • Video-MME (w/o subtitle)

    Inkling—
    Qwen3.5 397B83.7%
    Source

    Not directly comparable

  • MathVision

    Inkling—
    Qwen3.5 397B88.6%
    Source

    Not directly comparable

  • We-Math

    Inkling—
    Qwen3.5 397B87.9%
    Source

    Not directly comparable

  • DynaMath

    Inkling—
    Qwen3.5 397B86.3%
    Source

    Not directly comparable

  • SimpleVQA

    Inkling—
    Qwen3.5 397B67.1%
    Source

    Not directly comparable

  • MMLongBench-Doc

    Inkling—
    Qwen3.5 397B61.5%
    Source

    Not directly comparable

  • CC-OCR

    Inkling—
    Qwen3.5 397B82.0%
    Source

    Not directly comparable

  • AI2D_TEST

    Inkling—
    Qwen3.5 397B93.9%
    Source

    Not directly comparable

  • CountBench

    Inkling—
    Qwen3.5 397B97.2%
    Source

    Not directly comparable

  • RefCOCO (avg)

    Inkling—
    Qwen3.5 397B92.3%
    Source

    Not directly comparable

  • ODINW13

    Inkling—
    Qwen3.5 397B47.0%
    Source

    Not directly comparable

  • ERQA

    Inkling—
    Qwen3.5 397B67.5%
    Source

    Not directly comparable

  • VideoMMMU

    Inkling—
    Qwen3.5 397B84.7%
    Source

    Not directly comparable

  • MLVU (M-Avg)

    Inkling—
    Qwen3.5 397B86.7%
    Source

    Not directly comparable

  • ScreenSpot Pro

    Inkling—
    Qwen3.5 397B65.6%
    Source

    Not directly comparable

  • V*

    Inkling—
    Qwen3.5 397B95.8%
    Source

    Not directly comparable

Knowledge

  • GPQA

    Inkling87.9%
    Source
    Qwen3.5 397B88.4%
    Source

    Qwen3.5 397B leads this result

  • GPQA-D

    Inkling87.9%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • HLE

    Inkling46%
    Source
    Qwen3.5 397B28.7%
    Source

    Inkling leads this result

  • HLE w/o tools

    Inkling30%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • GPQA Diamond (Vals)

    Inkling87.1%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • MMLU-Pro (Vals)

    Inkling86.3%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • SuperGPQA

    Inkling—
    Qwen3.5 397B70.4%
    Source

    Not directly comparable

  • MMLU-Pro

    Inkling—
    Qwen3.5 397B87.8%
    Source

    Not directly comparable

  • MMLU-Redux

    Inkling—
    Qwen3.5 397B94.9%
    Source

    Not directly comparable

  • C-Eval

    Inkling—
    Qwen3.5 397B93%
    Source

    Not directly comparable

Multilingual

  • MMLU-ProX

    Inkling—
    Qwen3.5 397B84.7%
    Source

    Not directly comparable

  • NOVA-63

    Inkling—
    Qwen3.5 397B59.1%
    Source

    Not directly comparable

  • INCLUDE

    Inkling—
    Qwen3.5 397B85.6%
    Source

    Not directly comparable

  • PolyMath

    Inkling—
    Qwen3.5 397B73.3%
    Source

    Not directly comparable

  • MAXIFE

    Inkling—
    Qwen3.5 397B88.2%
    Source

    Not directly comparable

Instruction following

  • IFBench

    Inkling79.8%
    Source
    Qwen3.5 397B76.5%
    Source

    Inkling leads this result

  • IFEval

    Inkling—
    Qwen3.5 397B92.6%
    Source

    Not directly comparable

Math

  • AIME26

    Inkling97.1%
    Source
    Qwen3.5 397B93.3%
    Source

    Inkling leads this result

  • FrontierMath v2 (Tiers 1-3)

    Inkling33.333%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • FrontierMath v2 (Tier 4)

    Inkling4.878%
    Source
    Qwen3.5 397B—

    Not directly comparable

  • HMMT Feb 2025

    Inkling—
    Qwen3.5 397B94.8%
    Source

    Not directly comparable

  • HMMT Nov 2025

    Inkling—
    Qwen3.5 397B92.7%
    Source

    Not directly comparable

  • HMMT Feb 2026

    Inkling—
    Qwen3.5 397B87.9%
    Source

    Not directly comparable

  • MMAnswerBench

    Inkling—
    Qwen3.5 397B80.9%
    Source

    Not directly comparable

76 public results · 10 shared

Related comparisons

Compare API pricingRead the methodologyOpen the model selector

Watch Inkling vs Qwen3.5 397B

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

Read a sample issue

Join 5,500+ readers.

Last updated October 14, 2026