Model comparison

GPT-5.6 Luna vs Qwen 3.8 Flash Next

OpenAI against Qwen, compared on context, price, and verified benchmark results.

Catalog record checked September 26, 2026Individual provider fields may changeEvidence confidence: High — see receipts below

OpenAI

GPT-5.6 Luna

Budget

Qwen

Qwen 3.8 Flash Next

Budget · Open weights

AI model capability comparison
Specification	GPT-5.6 Luna	Qwen 3.8 Flash Next
Provider
Provider	OpenAI	Qwen
Tier
Tier	Budget	Budget
Context window
Context window	Winner: 1.05M	262K
Max output
Max output	128K	Not verifiedUnverified
Input / 1M tokens
Input / 1M tokens	$0.20	Winner: $0.15
Output / 1M tokens
Output / 1M tokens	$1.20	Winner: $0.47
Weights
Weights	Closed	Open
Parameters
Parameters	previous-generation GPT-5.6 small tier	125B total / 6B active (MoE) + 51B n-gram embedding table
Reasoning levels
Reasoning levels	none, low, medium, high, xhigh, max	Not verifiedUnverified
Modalities
Modalities	text, image	text, image, video
License
License	Not disclosedUnverified	qwen-community-1.0
API model id
API model id	gpt-5.6-luna	qwen3.8-flash
Released
Released	July 9, 2026	August 26, 2026
Artificial Analysis Intelligence Index [max] (2026-09-26)
Artificial Analysis Intelligence Index [max] (2026-09-26)	37.3	Not verifiedUnverified
Artificial Analysis Intelligence Index (2026-09-26)
Artificial Analysis Intelligence Index (2026-09-26)	Not verifiedUnverified	39.8
SWE-bench Verified (2026-09-01)
SWE-bench Verified (2026-09-01)	93	Not verifiedUnverified
Terminal-Bench 2.1 (2026-07-09)
Terminal-Bench 2.1 (2026-07-09)	84.7	Not verifiedUnverified
LiveBench (2026-09-05)
LiveBench (2026-09-05)	73.6	Not verifiedUnverified

Prices are USD per million tokens at standard rates, excluding batch and caching discounts. Bold indicates the better figure where one is objectively better. 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.

2026-09-26
Artificial Analysis
Composite index blending reasoning, knowledge, and coding evals into one 0–100 score.
Comparable with caveatOnly same-version scores are comparable. v4.3.2 was rebaselined (the top score fell from 66 on v4.1.1 to ~58), so v4.1.1 figures must not be compared with v4.3.2 figures — check measuredAt.
2026-09-05
LiveBench official leaderboard (benchmark-owned), max effort
Contamination-resistant general capability across reasoning, coding, math, data analysis, and language, with monthly question refreshes.
Comparable with caveatRolling question set: observations months apart measure different question mixes. Record the measurement date and compare within ~1 month windows.
2026-09-01
vals.ai SWE-bench Verified leaderboard (independent, mini-swe-agent bash-only) (independent; OpenAI did not report SWE-bench Verified for the GPT-5.6 family)
Real GitHub issue resolution: does the model's patch pass the hidden tests.
Comparable with caveatPost-audit vendor claims and pre-audit scores sit on different task trust levels; scaffolding (agent harness, compute budget) also dominates results. Never aggregate across scaffolds.
2026-07-09
OpenAI GPT-5.6 announcement (vendor; chart transcribed by Vellum)
Agentic terminal work: multi-step tasks executed in a sandboxed shell environment.
Comparable with caveatNot comparable with Terminal-Bench 3.0 or 4.0 (different task sets) or v1; harness configuration (container, time limits) also shifts results.

Pricing tiers

GPT-5.6 Luna: Launch list was $1/$6; cut to $0.20/$1.20 on 2026-07-30. Above 272K input the whole request bills 2x input / 1.5x output. Cached input $0.02/MTok; cache writes 1.25x input; batch $0.10/$0.60.

Qwen 3.8 Flash Next: Hosted qwen3.8-flash lists $0.15/$0.47 per MTok on Alibaba Cloud Model Studio with a 1M context. The open checkpoint is 262,144 tokens native, extensible to 1M.

BudgetRecord checked September 26, 2026

GPT-5.6 Luna

GPT-5.6 Luna was OpenAI's cheapest GPT-5.6 tier; GPT-6 Luna replaced it on 2026-09-22 at half the price ($0.10/$0.50).

Best for

Classification
High-volume chat
Latency-sensitive pipelines

Watch out

Superseded by GPT-6 Luna, which costs half as much and matches it on the Artificial Analysis index; still served, not deprecated.

BudgetOpen weightsRecord checked September 26, 2026

Qwen 3.8 Flash Next

Qwen 3.8 Flash Next is Alibaba's experimental preview of the Qwen 4 architecture — very low active parameters plus an unusual n-gram embedding component for cheap long-context.

Best for

Early testing of Qwen 4 architecture
Cheap high-throughput work
Multimodal input

Watch out

Experimental preview; qwen-community-1.0 license is more restrictive than MIT/Apache; benchmark claims are vendor-reported.

When the cheaper one wins

Qwen 3.8 Flash Next is cheaper on output at $0.47 per million tokens against $1.20 for GPT-5.6 Luna — about 2.6×. 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. These are standard-tier API rates, excluding batch and cache discounts.

Run the model picker

Evidence confidence: High

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.
3/5 core specs verified on both sides — Not published for at least one side: max output, reasoning levels.
1 shared named benchmark with differing scores — Measured on: Artificial Analysis Intelligence Index.
Verified within the last 90 days — Newest catalog check was 2 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.

GPT-5.6 Luna: OpenAI — GPT-5.6 (accessed 2026-08-29)
GPT-5.6 Luna: OpenAI API pricing (gpt-5.6-luna $0.20/$1.20) (accessed 2026-08-29)
GPT-5.6 Luna: OpenAI — Introducing GPT-6 Sol and Luna (GPT-5.6 Luna successor) (accessed 2026-09-26)
Qwen 3.8 Flash Next: The New Stack — Qwen3.8-Flash previews Qwen4 (accessed 2026-08-29)
Qwen 3.8 Flash Next: Yotta Labs — Qwen 3.8-Flash-Next specs (accessed 2026-08-29)
Qwen 3.8 Flash Next: Hugging Face — Qwen/Qwen3.8-Flash-Next (accessed 2026-09-26)
Qwen 3.8 Flash Next: Alibaba Cloud Model Studio pricing (qwen3.8-flash $0.15/$0.47) (accessed 2026-09-26)

Related comparisons

Diving deeper on one model? GPT-5.6 Luna · Qwen 3.8 Flash Next

Common questions

GPT-5.6 Luna vs Qwen 3.8 Flash Next

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

Is GPT-5.6 Luna or Qwen 3.8 Flash Next cheaper for input?

Qwen 3.8 Flash Next is cheaper at $0.15 per million input tokens, against $0.20 for GPT-5.6 Luna — roughly 1.3× the price. Output tokens usually dominate a real bill, so weigh the output rate more heavily than the input rate. These are base rates; GPT-5.6 Luna has tiered pricing: Launch list was $1/$6; cut to $0.20/$1.20 on 2026-07-30. Above 272K input the whole request bills 2x input / 1.5x output. Cached input $0.02/MTok; cache writes 1.25x input; batch $0.10/$0.60. Qwen 3.8 Flash Next has tiered pricing: Hosted qwen3.8-flash lists $0.15/$0.47 per MTok on Alibaba Cloud Model Studio with a 1M context. The open checkpoint is 262,144 tokens native, extensible to 1M.

Is GPT-5.6 Luna or Qwen 3.8 Flash Next cheaper for output?

Qwen 3.8 Flash Next is cheaper at $0.47 per million output tokens, against $1.20 for GPT-5.6 Luna — roughly 2.6× the price. Output tokens usually dominate a real bill, so weigh the output rate more heavily than the input rate. These are base rates; GPT-5.6 Luna has tiered pricing: Launch list was $1/$6; cut to $0.20/$1.20 on 2026-07-30. Above 272K input the whole request bills 2x input / 1.5x output. Cached input $0.02/MTok; cache writes 1.25x input; batch $0.10/$0.60. Qwen 3.8 Flash Next has tiered pricing: Hosted qwen3.8-flash lists $0.15/$0.47 per MTok on Alibaba Cloud Model Studio with a 1M context. The open checkpoint is 262,144 tokens native, extensible to 1M.

Which has the larger context window, GPT-5.6 Luna or Qwen 3.8 Flash Next?

GPT-5.6 Luna accepts 1.05M tokens against 262K for Qwen 3.8 Flash Next. This only matters if you routinely send very long documents or large codebases.

Should I use GPT-5.6 Luna or Qwen 3.8 Flash Next?

Both sit in the budget tier, so the choice usually comes down to price and context rather than capability. GPT-5.6 Luna suits classification; Qwen 3.8 Flash Next suits early testing of qwen 4 architecture.

Can I self-host GPT-5.6 Luna or Qwen 3.8 Flash Next?

Qwen 3.8 Flash Next publishes open weights, but self-hosting and third-party provider access depend on the licence, hardware, serving support, and availability. GPT-5.6 Luna 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 Flash Next costs $0.15 per 1M tokens versus $0.20 for GPT-5.6 Luna — a 1.3x difference at the headline tier.
Context: GPT-5.6 Luna takes 1.05M against 262K for Qwen 3.8 Flash Next — only decisive if your prompts approach the smaller window.
Measured capability: Qwen 3.8 Flash Next leads Artificial Analysis Intelligence Index 39.8 to 37.3 (measured 2026-09-26).
Deployment: Qwen 3.8 Flash Next publishes weights you can self-host; the other is API-only.

Run the model picker Estimate stack cost All model comparisons

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