Skip to main content
AI Choice Engine

Model comparison

Kimi K3 vs Qwen 3.8 Max

Moonshot AI against Qwen, compared on context, price, and attributed benchmark results.

Recorded profile verification October 2, 2026Individual provider fields may changeEvidence confidence: Medium — see receipts below

Moonshot AI

Kimi K3

Frontier · Open weights

vs

Qwen

Qwen 3.8 Max

Frontier

AI model capability comparison
SpecificationKimi K3Qwen 3.8 Max
ProviderMoonshot AIQwen
TierFrontierFrontier
Context windowWinner: 1.05M1M
Max outputWinner: 1.05M131K
Input / 1M tokens$3Winner: $2
Output / 1M tokens$15Winner: $6
WeightsOpenClosed
Parameters2.8T total / 104B active (MoE)2.4T total (MoE; managed active count not verified)
Reasoning levelslow, high, maxlow, medium, xhigh
Modalitiestext, image, videotext, image, video
LicenseKimi K3 License (custom)Not verifiedUnverified
API model idkimi-k3qwen3.8-max
ReleasedJuly 16, 2026August 3, 2026
Artificial Analysis Intelligence Index [max] · source date Kimi K3: 2026-10-0144Not verifiedUnverified
Artificial Analysis Intelligence Index · source date Qwen 3.8 Max: 2026-10-01Not verifiedUnverified45
Terminal-Bench 2.1 [max] · source date Kimi K3: 2026-10-0188.3Not verifiedUnverified
Terminal-Bench 2.1 · source date Qwen 3.8 Max: 2026-10-01Not verifiedUnverified86.6
GPQA Diamond [max] · source date Kimi K3: 2026-10-0193.5Not verifiedUnverified
GPQA Diamond · source date Qwen 3.8 Max: 2026-10-01Not verifiedUnverified92.6
Humanity's Last Exam [max] · source date Kimi K3: 2026-10-0156Not verifiedUnverified
Humanity's Last Exam · source date Qwen 3.8 Max: 2026-08-14Not verifiedUnverified56.2
SWE-bench Verified · source date Kimi K3: 2026-10-0193.4Not verifiedUnverified

Dates beside benchmark scores refer to the linked source’s publication or capture date unless that source explicitly states when it measured the result. A shared URL alone does not establish comparable runs; unverified setups are marked. Prices are USD per million tokens at standard rates, excluding batch and caching discounts. Values we could not confirm from the provider are shown as “Not verified” rather than estimated.

Benchmark receipts

Where each score comes from, and how far it can be compared across models.

Pricing tiers

Kimi K3: $3/$15 per million input/output tokens; cached input $0.30. Cache writes add $3/MTok for 5-minute TTL or $6/MTok for 1-hour TTL; hits refresh TTL without another write charge. API access requires a successful top-up of at least $1. max_completion_tokens defaults to 131,072 and can be set up to 1,048,576. Kimi K3 uses a custom license; self-hosting has compute costs.

Qwen 3.8 Max: Model Studio International $2/$6 per MTok; implicit cache $0.25, explicit creation $2.50/read $0.17. China/Global $1.65/$4.951, cache $0.206, creation $2.063/read $0.137. Output 131,072 in both modes; input 991,808 non-thinking or 983,616 thinking, 1M context and 262,144 reasoning. Batch supported in China, unavailable in International/Global.

FrontierOpen weightsRecord checked October 1, 2026

Kimi K3

Kimi K3 is Moonshot AI's open-weight multimodal 2.8T/104B MoE model with a 1,048,576-token context.

Best for

  • Open-weight frontier work
  • Long-context
  • Multimodal

Watch out

Thinking is always on; preserve complete assistant reasoning/tool history. Vision input requires base64 or ms:// rather than public image URLs, and temperature/top_p/n/penalties are fixed. Current docs discourage production web-search use. The custom licence requires an agreement for Model-as-a-Service businesses above $20M aggregate licensee/affiliate revenue over consecutive 12 months; commercial products above 100M monthly active users or $20M monthly revenue require Kimi K3 UI attribution. Internal use and official/certified serving routes have separate exemptions. July 16 is Kimi Code's release date, distinct from weights publication.

FrontierRecord checked October 2, 2026

Qwen 3.8 Max

Qwen 3.8 Max is Alibaba's managed multimodal flagship, with a vendor-described 2.4T MoE scale and 1M context.

Best for

  • Hosted agentic work
  • Long-context
  • Multimodal

Watch out

The text-only open checkpoint has 262,144 native context with extension and requires thinking; its 95B active count is not verified for the managed API. Managed Max accepts visual input and supports thinking on/off, xhigh default and low/medium/xhigh controls, with preserved thinking by default. API reasoning_effort and thinking_budget cannot both be set. Regional web-search and Batch support differ. Independent AA 0902 results identify a snapshot, not every alias version.

When the cheaper one wins

Qwen 3.8 Max is cheaper on output at $6 per million tokens against $15 for Kimi K3 — about 2.5×. Use the cheaper tier for classification, extraction, summarisation, and any task where the expensive model’s extra score does not change the accepted output. The expensive one only pays if your hardest task actually fails on the cheap tier. On DeepSWE 1.1, Qwen 3.8 Max is 57.5% Pass@1 at $3.73/task versus Kimi K3 at 68.5% / $4.65/task. This compares standard short-context token rates, not total cost per accepted task. Include input, cache, context tiers, schedules, retries and review in your pilot.

Run the model picker

Evidence confidence: Medium

How strong and complete the evidence behind this comparison is — not a prediction of which model is better.

  • Pricing verified on both sides — Input and output rates are verified for both models.
  • 5/5 core specs verified on both sides — All core specifications verified for both models.
  • No directly aligned benchmark setup — No benchmark has matching cited sources, source dates, units, effort and recorded run details on both models.
  • Verified within the last 90 days — Older of the two catalog checks was 25 days ago.
  • Both models carry source citations — Each side has at least two catalog sources on record.

Source receipts

Each catalog figure was checked against the provider or an independent second source on the date shown.

Related comparisons

Diving deeper on one model? Kimi K3 · Qwen 3.8 Max

Common questions

Kimi K3 vs Qwen 3.8 Max

Answered from the verified figures on this page rather than general guidance.

Is Kimi K3 or Qwen 3.8 Max cheaper for input?
Qwen 3.8 Max is cheaper at $2 per million input tokens, against $3 for Kimi K3 — roughly 1.5× the price. Estimate both input and output token volume for your workload. These are base rates; Kimi K3: $3/$15 per million input/output tokens; cached input $0.30. Cache writes add $3/MTok for 5-minute TTL or $6/MTok for 1-hour TTL; hits refresh TTL without another write charge. API access requires a successful top-up of at least $1. max_completion_tokens defaults to 131,072 and can be set up to 1,048,576. Kimi K3 uses a custom license; self-hosting has compute costs. Qwen 3.8 Max: Model Studio International $2/$6 per MTok; implicit cache $0.25, explicit creation $2.50/read $0.17. China/Global $1.65/$4.951, cache $0.206, creation $2.063/read $0.137. Output 131,072 in both modes; input 991,808 non-thinking or 983,616 thinking, 1M context and 262,144 reasoning. Batch supported in China, unavailable in International/Global.
Is Kimi K3 or Qwen 3.8 Max cheaper for output?
Qwen 3.8 Max is cheaper at $6 per million output tokens, against $15 for Kimi K3 — roughly 2.5× the price. Estimate both input and output token volume for your workload. These are base rates; Kimi K3: $3/$15 per million input/output tokens; cached input $0.30. Cache writes add $3/MTok for 5-minute TTL or $6/MTok for 1-hour TTL; hits refresh TTL without another write charge. API access requires a successful top-up of at least $1. max_completion_tokens defaults to 131,072 and can be set up to 1,048,576. Kimi K3 uses a custom license; self-hosting has compute costs. Qwen 3.8 Max: Model Studio International $2/$6 per MTok; implicit cache $0.25, explicit creation $2.50/read $0.17. China/Global $1.65/$4.951, cache $0.206, creation $2.063/read $0.137. Output 131,072 in both modes; input 991,808 non-thinking or 983,616 thinking, 1M context and 262,144 reasoning. Batch supported in China, unavailable in International/Global.
Which has the larger context window, Kimi K3 or Qwen 3.8 Max?
Kimi K3 accepts 1.05M tokens against 1M for Qwen 3.8 Max. This only matters if you routinely send very long documents or large codebases.
Do Kimi K3 and Qwen 3.8 Max support the same reasoning levels?
Kimi K3 exposes low, high, max, while Qwen 3.8 Max exposes low, medium, xhigh.
Should I use Kimi K3 or Qwen 3.8 Max?
Both are classified in our frontier tier; this editorial grouping does not imply equal capability. Kimi K3 is a candidate for open-weight frontier work; Qwen 3.8 Max is a candidate for hosted agentic work. Compare accuracy, tool reliability, latency and cost on representative tasks.
Can I self-host Kimi K3 or Qwen 3.8 Max?
Kimi K3 publishes open weights, but self-hosting and third-party provider access depend on the licence, hardware, serving support, and availability. Qwen 3.8 Max is a closed model whose supported access paths are controlled by its provider.

Next step

Choosing between them

The verified figures that separate this pair, computed from the catalog rather than restated boilerplate.

  • Input price: Qwen 3.8 Max costs $2 per 1M tokens versus $3 for Kimi K3 — a 1.5x difference at the headline tier.
  • Context: Kimi K3 takes 1.05M against 1M for Qwen 3.8 Max — only decisive if your prompts approach the smaller window.
  • Deployment: Kimi K3 publishes weights you can self-host; the other is API-only.
Run the model pickerEstimate stack costAll model comparisons

Choosing a harness rather than a model — Cursor, Copilot, Claude Code, Windsurf, or Lovable? Compare agentic harnesses · Latest releases.