LLM Pricing
ModelsLeaderboardsToolsProvidersReleasesArticles

© 2026 LLM Pricing

About
·Contact
·Privacy
·RSS
← Model list

Qwen3.8 Max Preview

alibaba·alibaba/qwen3.8-max-preview·GA·Closed·qwen series
Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows

Specs & pricing

Input / output per 1M tokens
Official price
—
Lowest paid·AIHubMixGateway
$0.34 / $1.01
Blended $0.51 · 7.4× spread
Context
1,000,000
Output limit
131,072
Knowledge cutoff
—
Released / updated
2026-07-19 / 2026-07-19
Capabilities
✓ Reasoning✓ Tool use✓ Structured output✓ Temperature✓ Attachments
Modalities
TextImageVideo

Qwen3.8 Max Preview by alibaba is offered by 8 providers on this page. Public prices are shown for 6 of them. No official API price is listed for it; the prices below come from third-party channels. The lowest paid channel is AIHubMix at $0.338 / $1.01 per 1M. That channel is a third-party gateway, so confirm its availability and rate limits before depending on it. Prices vary about 7.4× across the providers that host it. It also has 2 covered by a paid subscription; free tiers usually carry rate limits, and subscription-covered access bills $0 per token only after the subscription fee.

The context window is 1,000,000 tokens, with an output limit of 131,072 tokens. It supports reasoning, tool use, and structured output. Accepted input modalities are Text, Image, and Video.

Available at 8 providers6 with public prices · 2 subscription-covered

  • ProviderInput / Output
  • Alibaba Token PlanGateway
    Subscription
    Cache read $0Context 1,000,000Output limit 131,072deprecated
  • Alibaba Token Plan (China)Gateway
    Subscription
    Cache read $0Context 1,000,000Output limit 131,072deprecated
  • AIHubMixGateway
    $0.34 / $1.01
    Cache read $0.068Context 1,000,000Output limit 131,072deprecated
  • DevPass (LLM Gateway)Gateway
    qwen3.8-max
    $1.54 / $4.63
    Cache read $0.20Context 1,000,000Output limit 1,000,000
  • Charm HyperGateway
    qwen3.8-max
    $2.00 / $6.00
    Cache read $0.25Context 1,000,000Output limit 65,536
  • Vercel AI GatewayCloud
    alibaba/qwen3.8-max
    $2.00 / $6.00
    Cache read $0.25Context 1,000,000Output limit 128,000
  • Venice AIGateway
    qwen-3-8-max
    $2.50 / $7.50
    Cache read $0.31Context 1,000,000Output limit 131,072
  • ImpossiblGateway
    $2.50 / $7.50
    Context 1,000,000Output limit 131,072deprecated
ProviderTierInputOutputCache readCache writeContextOutput limitStatus
Alibaba Token PlanGatewaySubscription$0$01,000,000131,072deprecated
Alibaba Token Plan (China)GatewaySubscription$0$01,000,000131,072deprecated
AIHubMixGateway$0.34$1.01$0.068$0.421,000,000131,072deprecated
DevPass (LLM Gateway)
qwen3.8-max
Gateway$1.54$4.63$0.20$2.501,000,0001,000,000
Charm Hyper
qwen3.8-max
Gateway$2.00$6.00$0.25—1,000,00065,536
Vercel AI Gateway
alibaba/qwen3.8-max
Cloud$2.00$6.00$0.25$2.501,000,000128,000
Venice AI
qwen-3-8-max
Gateway$2.50$7.50$0.31$3.131,000,000131,072
ImpossiblGateway$2.50$7.50——1,000,000131,072deprecated

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.

Reasoning control

effort = loweffort = mediumeffort = xhighbudget_tokens ≥ 0 ≤ 262,144Toggle (on / off)effort = noneeffort = minimaleffort = high

Interleaved thinking (reasoning between tool calls) is declared by 3 of 8 providers.

1 / 8 providers expose no reasoning control (reasoning_options: []).

Your usage cost

Common scenarios
1AIHubMix$85.85
2DevPass (LLM Gateway)$378.51
3Charm Hyper$490.00
4Vercel AI Gateway$490.00
5Venice AI$612.50
6Impossibl$875.00

2 more channels offer $0 (Alibaba Token Plan, Alibaba Token Plan (China)); free tiers usually have rate limits and no SLA, excluded from ranking.

The cheapest paid channel is the only channel.

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$3.00$02026-08-052026-09-28
2026-09-28Input list —Output list —Min blended $0.51

Benchmark15 items

NameConditionsScoreMetricSource
Terminal-Benchvariant: xhigh · v2.186.6accuracySource ↗
SWE-Bench Proharness: Claude Code · variant: xhigh67.7resolve rateSource ↗
DeepSWEharness: Claude Code · variant: xhigh · v1.156.6resolve rateSource ↗
NL2Repoharness: Claude Code · variant: xhigh55.9resolve rateSource ↗
FrontierSWEharness: Claude Code · variant: xhigh73.5dominance scoreSource ↗
MLS-Bench-Liteharness: Claude Code · variant: xhigh41scoreSource ↗
AutomationBenchvariant: xhigh · dataset: 600-task public subset27.3pass@1Source ↗
Toolathlon Verifiedvariant: xhigh72.5pass@1Source ↗
WideSearchvariant: xhigh81.9F1Source ↗
Humanity's Last Examvariant: xhigh, with tools56.2accuracySource ↗
GPQA Diamondvariant: xhigh92.6accuracySource ↗
Humanity's Last Examvariant: xhigh, no tools43.6accuracySource ↗
IFBenchvariant: xhigh82.8scoreSource ↗
OSWorld-Verifiedvariant: xhigh86.1success rateSource ↗
MMMU Provariant: xhigh82.3accuracySource ↗

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—Qwen3.8 Max 0902same lab—Qwen 3.8 Max Primesame lab—
Data partly from models.dev (MIT)