Skip to main content

Model comparison

Inkling vs Qwen3.5 397B

Data verified

Head-to-head evidence from 10 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

Thinking Machines Lab
69/100
Margin
12.3pts
← winning
56.73/100
3 category wins3 category wins

Verified leaderboard positions: Inkling #17; Qwen3.5 397B #20

BenchAlign evidence: Inkling not scored; Qwen3.5 397B estimated. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Inkling and Qwen3.5 397B share 10 comparable benchmark results. 6 of 8 categories are comparable. 6 results are unique to Inkling; 46 to Qwen3.5 397B.

Updated July 15, 2026
Shared results
10
Inkling only
6
Qwen3.5 397B only
46
Comparable categories
6 / 8

Pick Inkling if you want the stronger benchmark profile. Qwen3.5 397B only becomes the better choice if instruction following is the priority or you want the cheaper token bill.

Confidence note. This is a partial-evidence comparison with 10 shared benchmark results across 5 evidence categories; 6 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Inkling is clearly ahead on the provisional aggregate, 69 to 59. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Inkling's sharpest advantage is in agentic, where it averages 69.4 against 56.5. The single biggest benchmark swing on the page is HLE, 46% to 28.7%. Qwen3.5 397B does hit back in instruction following, so the answer changes if that is the part of the workload you care about most.

Inkling is also the more expensive model on tokens at $1.87 input / $4.68 output per 1M tokens, versus $0.60 input / $3.60 output per 1M tokens for Qwen3.5 397B. Qwen3.5 397B is the reasoning model in the pair, while Inkling is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Inkling gives you the larger context window at 1M, compared with 128K for Qwen3.5 397B.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for Inkling and Qwen3.5 397B
CategoryInklingΔQwen3.5 397B
AgenticInkling69.4Margin 12.9Qwen3.5 397B56.5
Inst. FollowingInkling79.8Margin 12.8Qwen3.5 397B92.6
MathInkling97.1Margin 6.5Qwen3.5 397B90.6
KnowledgeInkling51.7Margin 5.2Qwen3.5 397B56.9
MultimodalInkling76.5Margin 3.1Qwen3.5 397B79.6
CodingInkling68.6Margin 2.1Qwen3.5 397B66.5
ReasoningInklingNot measuredMarginNo overlapQwen3.5 397B63.2
MultilingualInklingNot measuredMarginNo overlapQwen3.5 397B84.7

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · InklingB · Qwen3.5 397B
  1. HLE

    Knowledge
    Source ↗
    A 46%B 28.7%
    Winner: InklingΔ 17.3
    HLE: Inkling scored 46%; Qwen3.5 397B scored 28.7%. Inkling wins this benchmark.
  2. BrowseComp

    Agentic
    Source ↗
    A 77.1%B 62%
    Winner: InklingΔ 15.1
    BrowseComp: Inkling scored 77.1%; Qwen3.5 397B scored 62%. Inkling wins this benchmark.
  3. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 63.8%B 52.5%
    Winner: InklingΔ 11.3
    Terminal-Bench 2.0: Inkling scored 63.8%; Qwen3.5 397B scored 52.5%. Inkling wins this benchmark.
  4. MMMU-Pro

    Multimodal
    Source ↗
    A 73.5%B 79%
    Winner: Qwen3.5 397BΔ 5.5
    MMMU-Pro: Inkling scored 73.5%; Qwen3.5 397B scored 79%. Qwen3.5 397B wins this benchmark.
  5. AIME26

    Math
    Source ↗
    A 97.1%B 93.3%
    Winner: InklingΔ 3.8
    AIME26: Inkling scored 97.1%; Qwen3.5 397B scored 93.3%. Inkling wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricInklingQwen3.5 397BComparison
Input / output priceUSD per 1M tokensInkling$1.87 input / $4.68 outputQwen3.5 397B$0.6 input / $3.6 outputQwen3.5 397B has the lower combined listed price.
Generation speedtokens per secondInklingNot availableQwen3.5 397B96 tok/sA complete speed comparison is not available.
First-answer latencyseconds to first tokenInklingNot availableQwen3.5 397B2.44 sA complete latency comparison is not available.
Context windowmaximum listed tokensInkling1MQwen3.5 397B128KInkling lists the larger context window.

Benchmark Deep Dive

AgenticInkling wins
BenchmarkInklingQwen3.5 397BResult
Terminal-Bench 2.0Source 63.8%52.5%Inkling leads
BrowseCompSource 77.1%62%Inkling leads
MCP AtlasSource 74.1%46.1%Inkling leads
Design Arena Agentic Web DevSource 1258Not comparable
Claw-EvalSource 56.8%Not comparable
QwenClawBenchSource 51.8%Not comparable
τ³-bench resultsSource 68.4%Not comparable
VITA-BenchSource 43.7%Not comparable
DeepPlanningSource 37.6%Not comparable
ToolathlonSource 36.3%Not comparable
MCP-TasksSource 74.2%Not comparable
WideResearchSource 74.0%Not comparable
τ²-bench resultsSource 95.6%Not comparable
Gert LabsSource 46.76%Not comparable
ResearchClawBenchSource 14.2%Not comparable
AA Agentic IndexSource 19.9%Not comparable
APEX-Agents-AASource 15.3%Not comparable
GDPval-AASource 23.1%Not comparable
GDPval-AASource 962Not comparable
CodingInkling wins
BenchmarkInklingQwen3.5 397BResult
SWE-bench VerifiedSource 77.6%76.2%Inkling leads
SWE-bench ProSource 54.3%50.9%Inkling leads
Terminal-Bench 2.0Source 63.8%Not comparable
LiveCodeBench v6Source 83.6%Not comparable
Terminal-Bench HardSource 40.9%Not comparable
AA-SciCodeSource 42.0%Not comparable
AA Coding IndexSource 48.2%Not comparable
Reasoning
BenchmarkInklingQwen3.5 397BResult
LongBench v2Source 63.2%Not comparable
AI-NeedleSource 68.7%Not comparable
AA-LCRSource 65.7%Not comparable
CritPtSource 1.7%Not comparable
KnowledgeQwen3.5 397B wins
BenchmarkInklingQwen3.5 397BResult
GPQASource 87.9%88.4%Qwen3.5 397B leads
GPQA-DSource 87.9%Not comparable
HLESource 46%28.7%Inkling leads
HLE w/o toolsSource 30%Not comparable
SuperGPQASource 70.4%Not comparable
MMLU-ProSource 87.8%Not comparable
MMLU-ReduxSource 94.9%Not comparable
C-EvalSource 93%Not comparable
Artificial Analysis Intelligence IndexSource 33.7%Not comparable
AA-GPQA DiamondSource 89.3%Not comparable
AA-HLESource 27.3%Not comparable
AA-Omniscience IndexSource -29.8%Not comparable
AA-Omniscience AccuracySource 31.4%Not comparable
AA-Omniscience Hallucination RateSource 89.1%Not comparable
MathInkling wins
BenchmarkInklingQwen3.5 397BResult
AIME26Source 97.1%93.3%Inkling leads
HMMT Feb 2025Source 94.8%Not comparable
HMMT Nov 2025Source 92.7%Not comparable
HMMT Feb 2026Source 87.9%Not comparable
MMAnswerBenchSource 80.9%Not comparable
Multilingual
BenchmarkInklingQwen3.5 397BResult
MMLU-ProXSource 84.7%Not comparable
NOVA-63Source 59.1%Not comparable
MultimodalQwen3.5 397B wins
BenchmarkInklingQwen3.5 397BResult
MMMU-ProSource 73.5%79%Qwen3.5 397B leads
CharXivSource 82%80.8%Inkling leads
CharXiv w/o toolsSource 78.1%Not comparable
MathVisionSource 88.6%Not comparable
VideoMMMUSource 84.7%Not comparable
ScreenSpot ProSource 65.6%Not comparable
V*Source 95.8%Not comparable
AA-MMMU-ProSource 77.3%Not comparable
Inst. FollowingQwen3.5 397B wins
BenchmarkInklingQwen3.5 397BResult
IFBenchSource 79.8%Not comparable
IFEvalSource 92.6%Not comparable
AA-IFBenchSource 78.8%Not comparable
Frequently Asked Questions (7)

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

Inkling is ahead on BenchLM's provisional leaderboard, 69 to 59. The biggest single separator in this matchup is HLE, where the scores are 46% and 28.7%.

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

Qwen3.5 397B has the edge for knowledge tasks in this comparison, averaging 56.9 versus 51.7. Inside this category, HLE is the benchmark that creates the most daylight between them.

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

Inkling has the edge for coding in this comparison, averaging 68.6 versus 66.5. Inside this category, SWE-bench Pro is the benchmark that creates the most daylight between them.

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

Inkling has the edge for math in this comparison, averaging 97.1 versus 90.6. Inside this category, AIME26 is the benchmark that creates the most daylight between them.

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

Inkling has the edge for agentic tasks in this comparison, averaging 69.4 versus 56.5. Inside this category, MCP Atlas is the benchmark that creates the most daylight between them.

Which is better for multimodal and grounded tasks, Inkling or Qwen3.5 397B?

Qwen3.5 397B has the edge for multimodal and grounded tasks in this comparison, averaging 79.6 versus 76.5. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.

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

Qwen3.5 397B has the edge for instruction following in this comparison, averaging 92.6 versus 79.8. Inkling stays close enough that the answer can still flip depending on your workload.

Related Comparisons

Last updated: July 15, 2026

The AI models change fast. We track them for you.

A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.