Skip to content
ModelCap

Qwen3.5-9B

Current board #114Official publisher releaseOpen weights

Qwen·qwen/qwen3.5-9b

Qwen3.5-9B is ModelCap current-board rank #114 with Index 44.9 (Modeled · peer benchmarks), from the ModelCap Index as of 22 Sept 2026, 02:49 UTC.

Qwen3.5-9B listed OpenRouter input price is $0.10 per 1M tokens and listed output price is $0.15 per 1M tokens, from OpenRouter listed prices as of 22 Sept 2026, 02:49 UTC.

Qwen3.5-9B has a published context limit of 262K tokens in the ModelCap catalogue as of 22 Sept 2026, 02:49 UTC.

Qwen3.5-9B has 6 published OpenRouter serving endpoints, from OpenRouter as of 22 Sept 2026, 02:49 UTC.

Qwen3.5-9B weight access is open weights, from ModelCap classification of published repository metadata as of 22 Sept 2026, 02:49 UTC.

Snapshot facts · download the public dataset · methodology

ModelCap Index · Current board
44.9#114
Estimated · 40% support · score interval 30.8–59.0
No admitted evaluation publication timestamp yet
Market Gravity
48.7
Secondary market signal
Input
$0.10
per 1M tokens
Output
$0.15
per 1M tokens
Context
262K
tokens
Providers
6
endpoints
Gateway spend
latest · Vercel share

Providers

Uptime measured over the last 30 minutes
ProviderInputOutputContextQuantUptime
Darkbloomfp4Cheapest$0.08$0.13262Kfp499.85%
SiliconFlowfp8$0.10$0.15262Kfp898.79%
DeepInfrabf16$0.10$0.15262Kbf1699.49%
Venicefp8$0.10$0.15256Kfp8100.0%
Parasailbf16$0.10$0.25262Kbf1699.32%
Together$0.17$0.25262K99.67%

Hugging Face Inference Providers

Exact repository identity

Provider economics and telemetry published by the Hugging Face Router. These are operational signals, not benchmark results.

ProviderInputOutputContextTTFTThroughputFeatures
deepinfra
live
$0.10$0.15262K428 ms23.5 tok/stools · structured
ovhcloud
live
$0.12$0.18262K458 ms38.5 tok/stools · structured
together
live
$0.17$0.25262K356 ms112.9 tok/stools · structured
featherless-ai
live

Reported evaluation leads (35)

Not independent measurement

Structured rows reported in this exact model repository at the pinned revision below. They are the publisher talking about itself: never measured evidence, never a score input on their own. Rows marked as used were laddered against independently measured peers inside the launch-card channel of this model's modeled placement, with the replay-measured optimism haircut applied; the others played no part.

Dataset / taskMetricValueRevisionProvenance
Delores-Lin/MDPBench
overall
65.7
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
digital
74.8
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
photographed
62.7
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
latin
72.5
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
de
72.8
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
en
72
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
es
72
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
fr
64.4
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
id
66.2
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
it
77.6
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
nl
74.5
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
pt
79.1
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
vi
74
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
non_latin
58.2
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
ar
53.4
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
hi
56.2
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
jp
55.7
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
ko
60.3
Unpinned
not published
Reported
Not used in launch estimate
Delores-Lin/MDPBench
ru
54.7
Unpinned
not published
Reported
Not used in launch estimate
Idavidrein/gpqa
diamond
81.7
Unpinned
not published
Reported
Not used in launch estimate

Showing 20 of 35 rows.

ModelCap Index · Current board

Methodology
44.9
Modeled · peer benchmarks · calibrated architecture-and-release-era fallback from static public metadata
Current board
#114
Public rank basis
Modeled · peer benchmarks (architecture)
Index support
40.0% · limited
Index interval
30.8–59.0
Rank posterior
Not available
Top 5 / Top 10 probability
Not available / Not available
Posterior as of
Snapshot timestamp unavailable
Evidence as of
No admitted evaluation publication timestamp
Identity binding
Exact catalogue product · qwen/qwen/qwen3.5-9b · aggregates disclosed benchmark configurations · endpoint configuration not claimed · artifact metadata Qwen/Qwen3.5-9B@c202236235762e1c871ad0ccb60c8ee5ba337b9a (not the evaluation revision)
Index observations used
0
Qualified Index evidence point
44.9
How this position was produced · fused launch placement

No general-family board lists this model yet. Its point estimate is the precision-weighted mean of every channel that spoke for it; each channel’s share is the weight it carried. Independent measurements replace this placement as boards list the model.

Fused estimate
44.9 ± 11.0
Independent share
0%
Shrink toward corpus
+2.7
Peer ceiling share
40%
Optimism haircut
−0.6 points · cross lab probe, 2 probes
Method
precision-weighted-v1
Channels fused into this model’s launch placement
ChannelEstimateShare
Publisher corpus prior · not independent
publisher-corpus-prior over 203 held-out anchors
53.5 ± 22.724%
Launch card against resolved peers · not independent
launch card against 3 resolved peers on 10 rows
42.2 ± 12.676%
  • · Launch card row aa-lcr: claims above every resolved peer
  • · Launch card row gpqa-diamond: claims above every resolved peer
  • · Launch card row hmmt-feb-25: inside the resolved peer range
  • · Launch card row hmmt-nov-25: inside the resolved peer range
  • · Launch card row ifbench: inside the resolved peer range
  • · Launch card row ifeval: claims above every resolved peer
  • · Launch card row livecodebench-v-6: claims below every resolved peer
  • · Launch card row mmlu-pro: inside the resolved peer range
  • · Launch card row mmlu-prox: claims above every resolved peer
  • · Launch card row mmlu-redux: inside the resolved peer range
Bayesian Evidence Score · measured evidence input only

BES normalizes admitted public benchmark evidence for measured Index rows. Its legacy score and observables prior do not define the public language rank.

Observed capability
Evidence status
No admitted measured evidence
General preference
Coding
Agents & tools
Reasoning
Evidence breadth
0%
Evidence coverage
0%
Market and catalogue signals · secondary, never the language Index rank
Usage (OpenRouter popularity)
21.5
Liquidity (providers × uptime)
81.6
Open reach (HF downloads)
93.9
Surface (context / tools / modalities)
90.3
Freshness
47.5
  • · OpenRouter popularity #93
  • · 6 live providers
  • · HF 30d downloads 9307599
  • · Catalogue: context 262144, reasoning, tools, modalities Text/Image/Video
  • · First seen 2026-03-10

These adoption and deployment observations remain context only. They do not change this model's ModelCap Index score or rank.

The bounded 0–100 capability score blends each source’s competitive placement with its published achievement, then combines capability families. Evidence breadth adds a modest uncertainty adjustment; price and popularity are not part of benchmark-led rank.

Experimental capability estimate

Estimate policy
49.5
33.5–65.5 uncertainty interval
Excluded from rank
Confidence
75%
Training support
209 models · 120 lineages
Out-of-domain check
in domain
Feature coverage
100%
Lineage-held-out error
10.0 MAE

Experimental estimate. This layer predicts from admitted evidence and safe metadata only. It never enters ModelCap Score or rank.

Deployment readiness

Metadata index
79.4
strong · 80% metadata coverage
Not capability
Artifact reproducibility
100.0
Access & legal clarity
100.0
Deployability
98.0
Evaluation provenance
0.0

Missing: evaluationProvenance.admitted-evaluation-observations, evaluationProvenance.capability-family-breadth, evaluationProvenance.evaluation-source-diversity, evaluationProvenance.identity-and-source-confidence, evaluationProvenance.evaluation-dataset-revisions.

A metadata completeness and deployability index—not a safety certification, quality grade, or production-readiness claim.

Market activity

Venue coverage
Market Gravity
48.7
OpenRouter weekly popularity
#93 of 322

Market Gravity uses OpenRouter’s full-catalogue popularity order. Sparse Vercel top-ten observations appear only when published and do not affect the score.

Arena preference

No matched Arena rating

This model is visible because it is available in the live catalogue, but it has no trustworthy community result yet. ModelCap does not infer a quality score from price, features, or market activity.

Market Gravity breakdown

Usage55%
21.5

OpenRouter weekly popularity across the full model catalogue.

Liquidity25%
81.6

Independent providers versus the model's open or closed cohort, weighted by uptime.

Open reach15%
93.9

Hugging Face 30-day downloads for open models; neutral for closed models.

Freshness5%
47.5

Time since first public availability, on a six-month half-life.

Specification

Context window
262K tokens
Max output
33K tokens
Inputs
Text, Image, Video
Outputs
Text
Tokenizer
Qwen3
Cached input tokens
Not offered
First seen on OpenRouter
10 Mar 2026

Capabilities

  • Supported: Reasoning
  • Supported: Tool use
  • Supported: Structured output
  • Supported: Response format
  • Not supported: Moderated

Weights & access

Open weights. Downloadable weights are published under an OSI or free-culture license in the current repository metadata. This describes weight availability, not whether the full training stack qualifies as Open Source AI.

Access
Open weights
Licence
apache-2.0
Pinned revision
c20223623576
Parameters
9.7B
Architecture
Qwen3_5ForConditionalGeneration
Model type
qwen3_5
Hugging Face downloads (30d)
9.3M
Downloads (all time)
63.7M
Likes
2K
Repository
Qwen/Qwen3.5-9B
Repository updated
2 Mar 2026, 00:51 UTC

Compare Qwen3.5-9B with

Pick any model

Benchmark leaderboards featuring Qwen3.5-9B

All leaderboards
Public benchmark boards on which Qwen3.5-9B has a published result
LeaderboardScoreSource rankConfigurationPublished
τ²-Bench Telecom (Artificial Analysis run)Artificial Analysis86.8%#87 of 436tau2:telecom:dual-control:pass-at-1:3-repeats:source-model="Qwen3.5 9B (Reasoning)":reasoning=true13 Sept 2026
Humanity's Last Exam (Artificial Analysis run)Artificial Analysis14.9%#218 of 615hle:may-2025:text-only-2158:no-tools:pass-at-1:source-model="Qwen3.5 9B (Reasoning)":reasoning=true13 Sept 2026

Alternatives to Qwen3.5-9B, by the numbers

Ranked neighbours

Models within 3 places of #114 on the ModelCap Index.

Cheaper at a similar or higher Index

Index score within 3 points of, or above, this model's, with a lower listed blended price (3:1 input:output).

Other ranked Qwen models

The same publisher's other current models with a public Index position.

Qwen3.5-9B: common questions

How does Qwen3.5-9B rank among AI models?

Qwen3.5-9B holds ModelCap Index position #114 of 248 ranked language models as of 22 September 2026, with an Index score of 44.9 (interval 30.8–59.0); evidence: modeled · peer benchmarks. The Index combines public benchmark boards with published uncertainty; the methodology page explains the weighting.

How much does Qwen3.5-9B cost per 1M tokens?

Qwen3.5-9B is listed at $0.10 per 1M input tokens and $0.15 per 1M output tokens as of 22 September 2026; the lowest output price among 6 providers is $0.13 on Darkbloom. Prices come from the OpenRouter catalogue and re-observe on every refresh. For 1,000 requests using 2,000 input and 500 output tokens each, the listed rates imply $0.275 total (2M input + 0.5M output). This excludes caching, batch discounts, prompt-length tiers, tool charges and retries; verify the chosen endpoint's rate and limits before budgeting.

What is the context window of Qwen3.5-9B?

Qwen3.5-9B has a published context window of 262K tokens (262,144) and a maximum output of 33K tokens. Individual providers can serve less than the published maximum; the providers table lists each endpoint's own limit.

Is Qwen3.5-9B open-weight?

Open weights (apache-2.0). Downloadable weights are published under an OSI or free-culture license in the current repository metadata. This describes weight availability, not whether the full training stack qualifies as Open Source AI. The repository is Qwen/Qwen3.5-9B on Hugging Face.

Which API providers serve Qwen3.5-9B?

6 providers serve Qwen3.5-9B through OpenRouter as of 22 September 2026: Darkbloom, SiliconFlow, DeepInfra, Venice, Parasail and Together. Each provider's price, context limit, quantization and measured uptime are in the providers table above.

Which benchmarks has Qwen3.5-9B been evaluated on?

Qwen3.5-9B has published results on 2 public boards tracked by ModelCap as of 22 September 2026: AA τ²-Bench Telecom 86.8% (#87 of 436) and AA Humanity's Last Exam 14.9% (#218 of 615). Each board page ranks every tracked model on that benchmark; the ModelCap Index combines them with published uncertainty.

How does Qwen3.5-9B compare with ContextPilot E4B?

Qwen3.5-9B ranks #114 (Index 44.9) and ContextPilot E4B ranks #115 (Index 44.6) on the ModelCap Index as of 22 September 2026. The head-to-head page lines up their benchmarks, listed prices, context windows and provider counts side by side.

When was Qwen3.5-9B released?

Qwen3.5-9B first appeared in the catalogue on 10 Mar 2026. It is the current version in its family. Rank and price movements since then are recorded on the site's changes feed.