LLM Pricing
ModelsLeaderboardsToolsProvidersReleasesArticles

© 2026 LLM Pricing

About
·Contact
·Privacy
·RSS
← Model list

Qwen3-Coder 30B-A3B Instruct

alibaba·alibaba/qwen3-coder-30b-a3b-instruct·Deprecated·Open weights·qwen series
Smaller Qwen coder for efficient local agents and repo-level fixes

Specs & pricing

Input / output per 1M tokens
Official price·Alibaba
$0.45 / $2.25
Blended $0.90 · Cache read —
Lowest paid·Hugging FaceCloud
$0.07 / $0.26
Blended $0.12 · 6.4× spread
3 more $0 channels
Context
262,144
Output limit
65,536
Knowledge cutoff
2025-04
Released / updated
2025-07-31 / 2025-07-31
Capabilities
Reasoning✓ Tool use✓ Structured output✓ TemperatureAttachments
⚠ Providers report capability flags inconsistently
Modalities
Text
Weights
Hugging Face

Qwen3-Coder 30B-A3B Instruct by alibaba is offered by 19 providers on this page. Public prices are shown for 16 of them. Its official list price is $0.45 per 1M input tokens and $2.25 per 1M output tokens. The lowest paid channel is Hugging Face at $0.07 / $0.26 per 1M, about 7.7× below the list price. It also has 3 free ($0) channels; free tiers usually carry rate limits, and subscription-covered access bills $0 per token only after the subscription fee.

Median output speed is 232 tokens per second, with 1.02s to the first token.

The context window is 262,144 tokens at the reference host, but hosts report different limits, from 128,000 to 262,144, so the usable window depends on the provider you pick. It supports tool use and structured output. The weights are open, so it can also be self-hosted or served through a gateway of your choice. Providers report the capability flags inconsistently, so verify a specific feature against the host you plan to use. Its training knowledge cuts off at 2025-04.

Quality & performanceArtificial Analysis · Intelligence Index v5

Intelligence—
Value—
Output speed232 tok/s
TTFT1.02 s

Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.

Available at 19 providers16 with public prices · 3 free

  • ProviderInput / Output
  • PendraGateway
    qwen3-coder:30b
    Free
    Context 262,144Output limit 65,536
  • ModelScopeCloud
    Qwen/Qwen3-Coder-30B-A3B-Instruct
    Free
    Context 262,144Output limit 65,536
  • LMStudioCloud
    qwen/qwen3-coder-30b
    Free
    Context 262,144Output limit 65,536
  • Hugging FaceCloud
    Qwen/Qwen3-Coder-30B-A3B-Instruct
    $0.07 / $0.26
    Context 262,144Output limit 65,536
  • SiliconFlowCloud
    Qwen/Qwen3-Coder-30B-A3B-Instruct
    $0.07 / $0.28
    Context 262,000 ⚠Output limit 262,000
  • TokenWattGateway
    $0.07 / $0.28
    Context 131,072 ⚠Output limit 65,536
  • HeliconeGateway
    $0.10 / $0.30
    Context 262,144Output limit 262,144
  • NanoGPTGateway
    $0.10 / $0.40
    Cache read $0.05Context 128,000 ⚠Output limit 65,536
  • SiliconFlow (China)Cloud
    Qwen/Qwen3-Coder-30B-A3B-Instruct
    $0.10 / $0.42
    Context 262,000 ⚠Output limit 262,000
  • Amazon BedrockCloud
    qwen.qwen3-coder-30b-a3b-v1:0
    $0.15 / $0.60
    Context 262,144Output limit 131,072deprecated
  • Alibaba (China)OfficialFirst-party
    $0.22 / $0.86
    tiered >32K: $0.32 · >128K: $0.54
    Context 262,144Output limit 65,536deprecated
ProviderTierInputOutputCache readCache writeContextOutput limitStatus
Pendra
qwen3-coder:30b
GatewayFree——262,14465,536
ModelScope
Qwen/Qwen3-Coder-30B-A3B-Instruct
CloudFree——262,14465,536
LMStudio
qwen/qwen3-coder-30b
CloudFree——262,14465,536
Hugging Face
Qwen/Qwen3-Coder-30B-A3B-Instruct
Cloud$0.07$0.26——262,14465,536
SiliconFlow
Qwen/Qwen3-Coder-30B-A3B-Instruct
Cloud$0.07$0.28——262,000 ⚠262,000
TokenWattGateway$0.07$0.28——131,072 ⚠65,536
HeliconeGateway$0.10$0.30——262,144262,144
NanoGPTGateway$0.10$0.40$0.05—128,000 ⚠65,536
SiliconFlow (China)
Qwen/Qwen3-Coder-30B-A3B-Instruct
Cloud$0.10$0.42——262,000 ⚠262,000
Amazon Bedrock
qwen.qwen3-coder-30b-a3b-v1:0
Cloud$0.15$0.60——262,144131,072deprecated
Alibaba (China)OfficialFirst-party$0.22
tiered >32K: $0.32 · >128K: $0.54
$0.86——262,14465,536deprecated

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Your usage cost

Common scenarios
1Hugging Face$27.00
2SiliconFlow$28.00
3TokenWatt$28.00
4NanoGPT$34.00
5Helicone$35.00
6SiliconFlow (China)$41.85
13Alibaba (China) · Official$86.25

3 more channels offer $0 (Pendra, ModelScope, LMStudio); free tiers usually have rate limits and no SLA, excluded from ranking.

Switch to Hugging Face to save $59.25/mo (69%)

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$2.25$02026-08-052026-10-10
2026-10-10Input list $0.45Output list $2.25Min blended $0.12

Benchmark3 items

NameConditionsScoreMetricSource
Artificial Analysis Coding Index—19.4indexSource ↗
SciCode—27.8percent correctSource ↗
Terminal-Bench Hard—15.2success rateSource ↗

The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.

Related models

Qwen4 Maxsame series—Qwen 3.6 Plus Uncensoredsame series—Qwen 3.8 27B Uncensoredsame series—Codestral 25.08cheaper alternative$0.30 / $0.90Nova Litecheaper alternative$0.06 / $0.24Ministral 3 14Bcheaper alternative$0.20 / $0.20Codestral (latest)cheaper alternative$0.30 / $0.90Qwen3 8Bsame lab$0.18 / $0.70Qwen3-Coder 480B-A35B Instructsame lab$1.50 / $7.50
Data partly from models.dev (MIT) · Alibaba official docs ↗