Qwen
·Released on Apr 27, 2026Qwen3.6 35B A3B API Benchmarks, Pricing & Provider Data
Compare Qwen3.6 35B A3B with another model
Choose a model to open its comparison page.
LLM
Qwen3.6 35B A3B benchmark, API pricing, and provider data cover 38 API providers, with prices starting at $0.020/request. Qwen3.6 35B A3B free API options are available from 1 provider.
Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per token. It uses a hybrid sparse mixture-of-experts architectur...
Quality
#87of 100
44.0
LMSpeed score
Cost
#71of 180
$0.020/ 1M · 8:1 in:out
$0.0032 in · $0.017 out
Category Performance
Observed-capability estimates with uncertainty reported separately from independent benchmark families.
- Coverage
- 7 / 8
- Methodology
- V3.0
#1Reasoning52.4RatedGlobal rank #333/4 Measured dimensions
80% interval: 43.7–61.0
abstract logicPrior only
scientific causal49.6
critpt · critpt / gpqa · gpqa / hle · hle
multistep constraints56.4
mmlu_pro · mmlu_pro
evidence verification51.1
lcr · lcr
#2Instruction following50.6Provisional1/4 Measured dimensions
80% interval: 34.6–66.6
constraint followingPrior only
structured outputPrior only
novel instruction generalization50.6
ifbench · aa_if_bench
long multiturn instructionPrior only
#3Agents46RatedGlobal rank #444/4 Measured dimensions
80% interval: 40.1–51.9
planning46.7
deep_planning · gert_labs
tool use39.8
mcp_atlas · mcp_atlas / tau · tau2_bench / tau · tau3_bench / toolathlon · toolathlon
environment execution37.5
gdpval_aa · benchlm_agentic_gdpval_aa / terminalbench · benchlm_agentic_terminal_bench2 / vita_bench · vita_bench / wide_research · wide_research
recovery reliability59.9
claw_eval · claw_eval
#4Multimodal44.1RatedGlobal rank #94/4 Measured dimensions
80% interval: 37.0–51.1
perception ocr45.5
simplevqa · simple_vqa
document spatial46.2
charxiv · charxiv
visual reasoning44.9
mmmu_pro · mmmu_pro
video action39.6
video_mmmu · video_mmmu
#5Knowledge39.1Estimated2/4 Measured dimensions
80% interval: 26.9–51.3
broad knowledge38.2
c_eval · c_eval
professional knowledge40
supergpqa · super_gpqa
factualityPrior only
retrieval open bookPrior only
#6Coding37.7RatedGlobal rank #394/4 Measured dimensions
80% interval: 31.7–43.7
code generation38.8
livecodebench · livecodebench / scicode · scicode
repository engineering28.2
nl2repo · nl2_repo / swe_pro · swe_pro
debugging testing40.8
swe_multilingual · benchlm_coding_swe_multilingual / swe_verified · swe_verified
tooling quality43.1
terminalbench · benchlm_coding_terminal_bench2
#7Math34.6Estimated2/4 Measured dimensions
80% interval: 23.1–46.1
foundational mathPrior only
competition math33.9
aime · aime2026 / hmmt · hmmt_feb2025 / hmmt · hmmt_feb2026 / hmmt · hmmt_nov2025
proof frontierPrior only
applied tool math35.3
mmanswer · mm_answer_bench
No data:Multilingual
Specifications
Input and output token limits for this model, plus how it ranks on long-context understanding.
INPUT
262.1Ktokens
≈ 314.6 pages of text
OUTPUT
262.1Ktokens
8K128K1M4M
262.1K
Features
Technical Details
- Input
- Output
- Total parameters
- 35B
- Active parameters
- 3B
- Released
- Apr 2026
- Tokenizer
- Qwen
- Architecture
- text+image+video->text
- Moderated
- No
- Supported parameters
- frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p
Rankings
Excels at
It's decent at
Falls behind in
Detailed scores
Updated: Aug 22, 2026Overall
undefined metric} other undefined metrics}}
Overall score44.0#87 / 100
Overall
undefined metric} other undefined metrics}}
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Speed & latency
undefined metric} other undefined metrics}}
Output speed139.8 tok/s#20 / 74Time to first token1.17 s#33 / 74
Speed & latency
undefined metric} other undefined metrics}}
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Output speed
LMSpeed rank#20
Score139.8 tok/s
Sourceartificialanalysis.ai
UpdatedAug 22, 2026
Confidence4
Time to first token
LMSpeed rank#33
Score1.17 s
Sourceartificialanalysis.ai
UpdatedAug 22, 2026
Confidence4
Pricing
undefined metric} other undefined metrics}}
Input price$0.375/M#71 / 180Output price$2.25/M#80 / 180
Pricing
undefined metric} other undefined metrics}}
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Agents
V3.0undefined metric} other undefined metrics}} · Rated
Score46#4480% interval40.1–51.94/4 Measured dimensionsAgentic score36.9#56 / 69Terminal-Bench 2.051.5#43 / 53
Agents
V3.0undefined metric} other undefined metrics}} · Rated
Dimensions and evidence
Planning & decomposition46.7
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
deep_planning · gert_labs · z -0.42 · q 1.00
Tool use39.8
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
mcp_atlas · mcp_atlas · z -0.84 · q 1.00
tau · tau2_bench · z 0.65 · q 1.00
tau · tau3_bench · z -1.15 · q 1.00
toolathlon · toolathlon · z -2.32 · q 1.00
Environment & long-horizon execution37.5
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
gdpval_aa · benchlm_agentic_gdpval_aa · z -0.34 · q 1.00
terminalbench · benchlm_agentic_terminal_bench2 · z -0.75 · q 1.00
vita_bench · vita_bench · z 0.00 · q 1.00
wide_research · wide_research · z -3.00 · q 1.00
Recovery & completion reliability59.9
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
claw_eval · claw_eval · z 1.56 · q 1.00
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Coding
V3.0undefined metric} other undefined metrics}} · Rated
Score37.7#3980% interval31.7–43.74/4 Measured dimensionsLiveCodeBench80.4%#17 / 115SciCode35.8%#123 / 204
Coding
V3.0undefined metric} other undefined metrics}} · Rated
Dimensions and evidence
Code generation38.8
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
livecodebench · livecodebench · z 0.97 · q 1.00
scicode · scicode · z -3.00 · q 1.00
Repository engineering28.2
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
nl2repo · nl2_repo · z -2.74 · q 1.00
swe_pro · swe_pro · z -1.32 · q 1.00
Debugging & testing40.8
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
swe_multilingual · benchlm_coding_swe_multilingual · z -1.50 · q 1.00
swe_verified · swe_verified · z -0.64 · q 1.00
Tool-assisted development & quality43.1
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
terminalbench · benchlm_coding_terminal_bench2 · z -0.92 · q 1.00
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
SciCodeV3 evidence
LMSpeed rank#123
Score35.8%
Sourceartificialanalysis.ai
UpdatedAug 22, 2026
Confidence4
Reasoning
V3.0undefined metric} other undefined metrics}} · Rated
Score52.4#3380% interval43.7–61.03/4 Measured dimensionsMMLU-Pro85.2%#29 / 129GPQA84.1%#68 / 211
Reasoning
V3.0undefined metric} other undefined metrics}} · Rated
Dimensions and evidence
Abstract logicPrior only
Scientific & causal reasoning49.6
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
critpt · critpt · z -0.72 · q 1.00
gpqa · gpqa · z 0.27 · q 1.00
hle · hle · z 0.00 · q 1.00
Multi-step constraints56.4
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
mmlu_pro · mmlu_pro · z 0.65 · q 1.00
Evidence integration & verification51.1
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
lcr · lcr · z -0.21 · q 1.00
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Knowledge
V3.0undefined metric} other undefined metrics}} · Estimated
Score39.180% interval26.9–51.32/4 Measured dimensionsKnowledge score46.6#63 / 69SuperGPQA64.7#14 / 16
Knowledge
V3.0undefined metric} other undefined metrics}} · Estimated
Dimensions and evidence
Broad knowledge38.2
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
c_eval · c_eval · z -1.65 · q 0.63
Professional knowledge40
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
supergpqa · super_gpqa · z -1.08 · q 1.00
FactualityPrior only
Retrieval & open-book usePrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Artificial Analysis Intelligence Index
LMSpeed rank#68
Score32.1
Sourcebenchlm.ai
UpdatedAug 20, 2026
Confidence3
AA-Omniscience Hallucination Rate
LMSpeed rank#74
Score50.5
Sourcebenchlm.ai
UpdatedAug 20, 2026
Confidence3
Math
V3.0undefined metric} other undefined metrics}} · Estimated
Score34.680% interval23.1–46.12/4 Measured dimensionsHMMT Feb 202590.7#7 / 7HMMT Nov 202589.1#9 / 9
Math
V3.0undefined metric} other undefined metrics}} · Estimated
Dimensions and evidence
Foundational mathPrior only
Competition math33.9
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
aime · aime2026 · z -1.92 · q 1.00
hmmt · hmmt_feb2025 · z -1.50 · q 0.88
hmmt · hmmt_feb2026 · z -0.51 · q 1.00
hmmt · hmmt_nov2025 · z -1.44 · q 1.00
Advanced proofsPrior only
Applied & tool-assisted math35.3
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
mmanswer · mm_answer_bench · z -1.91 · q 1.00
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Multilingual
V3.0undefined metric} other undefined metrics}} · No data
No data80% interval30.8–69.20/4 Measured dimensions
Multilingual
V3.0undefined metric} other undefined metrics}} · No data
Dimensions and evidence
Cross-language understandingPrior only
Multilingual generationPrior only
Reasoning transferPrior only
Low-resource robustnessPrior only
Multimodal
V3.0undefined metric} other undefined metrics}} · Rated
Score44.1#980% interval37.0–51.14/4 Measured dimensionsMultimodal Grounded score52.1#34 / 44MMMU81.7#5 / 7
Multimodal
V3.0undefined metric} other undefined metrics}} · Rated
Dimensions and evidence
Perception & OCR45.5
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
simplevqa · simple_vqa · z -0.58 · q 1.00
Document & spatial understanding46.2
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
charxiv · charxiv · z -0.39 · q 1.00
Visual reasoning44.9
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
mmmu_pro · mmmu_pro · z -0.74 · q 1.00
Video & grounded action39.6
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
video_mmmu · video_mmmu · z -0.63 · q 1.00
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Instruction following
V3.0undefined metric} other undefined metrics}} · Provisional
Score50.680% interval34.6–66.61/4 Measured dimensionsAA-IFBench64.4#46 / 84
Instruction following
V3.0undefined metric} other undefined metrics}} · Provisional
Dimensions and evidence
Constraint followingPrior only
Structured outputPrior only
Novel-instruction generalization50.6
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
ifbench · aa_if_bench · z -0.19 · q 1.00
Multi-turn & long instructionsPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
OpenRouter endpoints
10 endpointsThird-party OpenRouter endpoint data, shown separately from LMSpeed measurements. Some 30-minute live performance fields only appear after syncing with an OpenRouter API key.
| Provider endpoint | Input | Output | 1d uptime | 30m latency | 30m throughput | Context / output |
|---|---|---|---|---|---|---|
Venice venice/fp8 | $0.100/M | $1/M | 99.9% | — | — | undefined tokens / undefined tokens |
Io Net io-net/fp8 | $0.190/M | $1.19/M | 99.6% | — | — | undefined tokens / undefined tokens |
AkashML akashml/fp8 | $0.140/M | $1/M | 99.5% | — | — | undefined tokens / undefined tokens |
CoreWeave coreweave/fp8 | $0.250/M | $1.25/M | 99.5% | — | — | undefined tokens / undefined tokens |
Parasail parasail/fp8 | $0.150/M | $1/M | 99.4% | — | — | undefined tokens / undefined tokens |
AtlasCloud atlas-cloud/fp8 | $0.186/M | $1.11/M | 99.2% | — | — | undefined tokens / undefined tokens |
DeepInfra deepinfra/fp8 | $0.100/M | $0.950/M | 97.3% | — | — | undefined tokens / undefined tokens |
SiliconFlow siliconflow/fp8 | $0.200/M | $1.60/M | 96.9% | — | — | undefined tokens / undefined tokens |
Phala phala | $0.200/M | $1.27/M | 90.4% | — | — | undefined tokens / undefined tokens |
Darkbloom darkbloom/fp4 | $0.080/M | $0.750/M | 85.2% | — | — | undefined tokens / undefined tokens |
Pricing Comparison
Compare Qwen3.6 35B A3B API pricing across 37 providers. Prices range from $0.020/request to $2.57/M. AAAI offers the lowest rate at $0.020/request. 1 provider offers free API credits or a free tier.
| Provider | Health | Model Variant | Group | Input ($/M) | Output ($/M) | Speed (t/s) | First token | Audit |
|---|---|---|---|---|---|---|---|---|
Moyanjdc API Free | L1 100% | qwen3.6-35b-a3b | default | Free | Free | — | — | — |
L1 99% | qwen3.6-35b-a3b | default | $0.881/M | $5.29/M | — | — | — | |
L1 99% | qwen3.6-35b-a3b | default | $2.57/M | $2.57/M | — | — | — | |
L1 99% | qwen3.6-35b-a3b | default | -39%$0.229/M | -39%$1.38/M | — | — | — | |
L1 99% | qwen3.6-35b-a3b | default | -39%$0.229/M | -39%$1.38/M | — | — | — | |
L1 53% | Qwen3.6-35B-A3B | default | $0.020/request | - | — | — | — | |
L1 100% | qwen3.6-35b-a3b | default | -67%$0.123/M | -67%$0.740/M | — | — | — | |
L1 100% | qwen3.6-35b-a3b | 企业用户 | $1.80/M | $18.00/M | — | — | — | |
L1 0% | qwen3.6-35b-a3b | 2api | $1.00/request | - | — | — | — |
Alternatives & Similar Models
GPT-5.4
gpt-5-4
OpenAI GPT-5.4 extends the GPT-5 family with stronger instruction following, deeper tool use, and improved performance on coding, math, and long-document analysis.
14 shared providers
GPT-5.3 Codex
gpt-5-3-codex
OpenAI GPT-5.3 Codex is a code-specialized variant in the GPT-5 series, optimized for code generation, debugging, and software development tasks.
14 shared providers
GPT-5.2
gpt-5-2
OpenAI GPT-5.2 is a GPT-5 series model emphasizing advanced reasoning, multimodal understanding, and high-quality outputs for complex enterprise workloads.
14 shared providers
GPT-5.4 Mini
gpt-5-4-mini
OpenAI GPT-5.4 Mini is a compact language model in the GPT-5 series, optimized for quick responses and high throughput.
14 shared providers
Claude Haiku 4.5
claude-haiku-4-5
Anthropic Claude Haiku 4.5 delivers fast, low-cost responses while retaining solid instruction following for chat, classification, and lightweight coding.
14 shared providers
GLM-5.1
glm-5-1
Zhipu GLM-5.1 is a next-generation GLM model aimed at frontier reasoning, coding, and bilingual agent applications.
14 shared providers
Frequently Asked Questions
- What benchmark data does Qwen3.6 35B A3B include?
- LMSpeed shows Qwen3.6 35B A3B benchmark context, API price, output speed, first-token latency, and provider data across 38 providers when those signals are available.
- What is the Qwen3.6 35B A3B API price?
- Qwen3.6 35B A3B has pricing from 38 providers, ranging from $0.020/request to $2.57/M. AAAI has the lowest listed price.
- What does the Qwen3.6 35B A3B API pricing table include?
- The Qwen3.6 35B A3B API pricing table compares 38 providers by input price, output price, free tier status, speed, first-token latency, and recent health data when available.
- Which provider has the cheapest Qwen3.6 35B A3B API pricing?
- AAAI currently has the lowest listed Qwen3.6 35B A3B price at $0.020/request across 38 providers.
- Is Qwen3.6 35B A3B API free?
- Yes, Qwen3.6 35B A3B free API options are available through 1 providerundefined other undefined} on LMSpeed, including Moyanjdc API. These providers offer free API credits or a free tier with no per-token charges.
- Where can I get Qwen3.6 35B A3B free API access?
- LMSpeed currently lists 1 free API providerundefined other undefined} for Qwen3.6 35B A3B: Moyanjdc API. Check each provider row before using it because free tier limits can change.