GLM 5.3 Flash vs GPT-6.2 Astra
GPT-6.2 Astra leads GLM 5.3 Flash on coding, 87.0 to 56.3 on the Vibe Coding Index, and GLM 5.3 Flash costs 80.0x less per output token ($0.50 vs $40.00 per million). Updated November 22, 2026; 56 benchmarks have a result for both.
GLM 5.3 Flash or GPT-6.2 Astra: which to pick
Pick GPT-6.2 Astra for coding: it scores 87.0 on the Vibe Coding Index against 56.3 for GLM 5.3 Flash, though on 4 benchmarks against 6.
Pick GLM 5.3 Flash for output tokens 80.0x cheaper ($0.50 vs $40.00 per million) and work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.
- Shared results
- 56
- GLM 5.3 Flash only
- 56
- GPT-6.2 Astra only
- 20
- Like-for-like categories
- 8 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
GPT-6.2 Astra
GPT-6.2 Astra scores 87.0 against 56.3.
6 and 4 benchmarks behind the scores
Intelligence
AA Intelligence Index
GPT-6.2 Astra
GPT-6.2 Astra leads, 62.1 against 27.8.
Same board, both measured
Reasoning
GPQA Diamond
No call
Only GLM 5.3 Flash has a GPQA Diamond result, 86.4%.
Directional only
Knowledge
MMLU Pro
No call
Only GLM 5.3 Flash has a MMLU Pro result, 86.1%.
Directional only
Web design
Design Arena: websites
GPT-6.2 Astra
GPT-6.2 Astra leads, 1333 against 1252.
Same board, both measured
General
Vals Index
GPT-6.2 Astra
GPT-6.2 Astra leads, 67.7% against 50.9%.
Same board, both measured
Long documents
Listed context window
GPT-6.2 Astra
GPT-6.2 Astra takes 1.05M tokens against 1.05M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
GLM 5.3 Flash
GLM 5.3 Flash $0.0004, GPT-6.2 Astra $0.028 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only GLM 5.3 Flash has the number, $0.02.
Directional only
At a glance
- Coding index
- GLM 5.3 Flash 56.3 from 6 benchmarks, GPT-6.2 Astra 87.0 from 4 benchmarks.
- Widest category
- Math, 42.0 points apart on average over 3 shared benchmarks.
- Widest single gap
- Mystery game puzzles: GLM 5.3 Flash 8.0% against GPT-6.2 Astra 95.0%, 87.0 points apart on a 0 to 100 scale.
- Output price
- GLM 5.3 Flash $0.50 against GPT-6.2 Astra $40.00 per million output tokens, 80.0x apart.
- Context window
- GLM 5.3 Flash takes 1.05M tokens, GPT-6.2 Astra 1.05M.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | GLM 5.3 Flash | Bars | GPT-6.2 Astra | Ties |
|---|---|---|---|---|
| Coding | 2 | 9 | ||
| Intelligence | 0 | 2 | ||
| Agentic work | 2 | 6 | ||
| Reasoning | 2 | 8 | ||
| Math | 0 | 3 | ||
| Knowledge | 3 | 1 | ||
| Web design | 0 | 6 | ||
| General | 10 | 2 | ||
| All shared boards | 19 | 37 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index29
| Benchmark | GLM 5.3 Flash | GPT-6.2 Astra | Leader |
|---|---|---|---|
| LiveBench coding | 70.9% | 91.9% | GPT-6.2 Astra |
| LMArena WebDev | 1610 | 1766 | GPT-6.2 Astra |
| LiveCodeBench (Vals AI) | 80.5% | Not measured | Only GLM 5.3 Flash measured |
| SWE-bench Verified (Vals AI) | 92.0% | Not measured | Only GLM 5.3 Flash measured |
| Terminal-Bench 4 (Vals AI) | 37.4% | 63.1% | GPT-6.2 Astra |
| Vibe Code Bench (Vals AI) | 30.8% | 89.9% | GPT-6.2 Astra |
| Code migration (Vals AI) | 20.5% | Not measured | Only GLM 5.3 Flash measured |
| ProgramBench (Vals AI) | 0.0% | Not measured | Only GLM 5.3 Flash measured |
| IOI (Vals AI) | 52.5% | Not measured | Only GLM 5.3 Flash measured |
| Terminal-Bench 2.1 (Vals AI) | 62.9% | Not measured | Only GLM 5.3 Flash measured |
| Vibe Code Bench 1-100 (Vals AI) | 16.0% | Not measured | Only GLM 5.3 Flash measured |
| SkillsBench (Vals AI) | 40.2% | Not measured | Only GLM 5.3 Flash measured |
| CyberBench (Vals AI) | 67.9% | Not measured | Only GLM 5.3 Flash measured |
| SciCode | 51.6% | Not measured | Only GLM 5.3 Flash measured |
| Terminal-Bench 4.0 (Artificial Analysis run) | 32.8% | 64.1% | GPT-6.2 Astra |
| FrontierCode 2.0 | 17.6% | 49.0% | GPT-6.2 Astra |
| FrontierCode | 31.8% | Not measured | Only GLM 5.3 Flash measured |
| FrontierSWE | 18.1% | Not measured | Only GLM 5.3 Flash measured |
| MirrorCode | Not measured | 67.8% | Only GPT-6.2 Astra measured |
| ALE-Bench | 303.6 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: coding | 1509 | 1450 | GLM 5.3 Flash |
| LMArena Text: software and IT | 1501 | 1443 | GLM 5.3 Flash |
| LMArena WebDev: React | 1630 | 1770 | GPT-6.2 Astra |
| LMArena WebDev: HTML | 1579 | 1854 | GPT-6.2 Astra |
| LMArena WebDev: from an image | 1581 | Not measured | Only GLM 5.3 Flash measured |
| GameBench 2 | 60.6 | 97.2 | GPT-6.2 Astra |
| Terminal-Bench 4.0 | 36.7% | Not measured | Only GLM 5.3 Flash measured |
| Neither model has a result on the other 25 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | GLM 5.3 Flash | GPT-6.2 Astra | Leader |
|---|---|---|---|
| AA Intelligence Index | 27.8 | 62.1 | GPT-6.2 Astra |
| Epoch Capabilities Index | 153.8 | 172.8 | GPT-6.2 Astra |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index26
| Benchmark | GLM 5.3 Flash | GPT-6.2 Astra | Leader |
|---|---|---|---|
| Design Arena: agents building web apps | Not measured | 1334 | Only GPT-6.2 Astra measured |
| Design Arena: agents building full-stack apps | Not measured | 1351 | Only GPT-6.2 Astra measured |
| Design Arena: agents building mobile apps | Not measured | 1305 | Only GPT-6.2 Astra measured |
| Design Arena: agents building Android apps | Not measured | 1257 | Only GPT-6.2 Astra measured |
| Design Arena: spreadsheets | Not measured | 1278 | Only GPT-6.2 Astra measured |
| LiveBench research | 73.9% | 91.5% | GPT-6.2 Astra |
| LiveBench browser | 59.8% | 93.0% | GPT-6.2 Astra |
| LiveBench long agentic | 20.7% | 70.8% | GPT-6.2 Astra |
| SREBench (Vals AI) | 0.8% | Not measured | Only GLM 5.3 Flash measured |
| Finance Agent (Vals AI) | 57.9% | Not measured | Only GLM 5.3 Flash measured |
| Tax Agent Bench (Vals AI) | 31.6% | 21.8% | GLM 5.3 Flash |
| Legal Research Bench (Vals AI) | 45.2% | 42.8% | GLM 5.3 Flash |
| Harvey's Legal Agent Benchmark (Vals AI) | 10.4% | 16.3% | GPT-6.2 Astra |
| Excel Modeling Benchmark (Vals AI) | 55.9% | Not measured | Only GLM 5.3 Flash measured |
| BioMysteryBench (Vals AI) | Not measured | 80.7% | Only GPT-6.2 Astra measured |
| Terminal-Bench-Science (Vals AI) | 2.9% | 71.4% | GPT-6.2 Astra |
| Vals Science Index | Not measured | 68.4% | Only GPT-6.2 Astra measured |
| Robo-bench (Vals AI) | Not measured | 32.7% | Only GPT-6.2 Astra measured |
| APEX-Agents | 52.8% | 72.2% | GPT-6.2 Astra |
| PostTrainBench | 36.8% | Not measured | Only GLM 5.3 Flash measured |
| SpireClimb | Not measured | 41.3% | Only GPT-6.2 Astra measured |
| LMArena Search | Not measured | 1191 | Only GPT-6.2 Astra measured |
| Toolathlon-Verified | 78.4% | Not measured | Only GLM 5.3 Flash measured |
| Neither model has a result on the other 21 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index28
| Benchmark | GLM 5.3 Flash | GPT-6.2 Astra | Leader |
|---|---|---|---|
| GPQA Diamond | 86.4% | Not measured | Only GLM 5.3 Flash measured |
| LiveBench reasoning | 77.6% | 91.6% | GPT-6.2 Astra |
| MysteryMechanism (Vals AI) | Not measured | 53.1% | Only GPT-6.2 Astra measured |
| MMMU Pro (Vals AI) | 86.0% | Not measured | Only GLM 5.3 Flash measured |
| GPQA Diamond (Artificial Analysis run) | 91.2% | Not measured | Only GLM 5.3 Flash measured |
| Humanity's Last Exam (Artificial Analysis run) | 39.9% | Not measured | Only GLM 5.3 Flash measured |
| HLE-Diamond (Artificial Analysis run) | 18.2% | 72.2% | GPT-6.2 Astra |
| CritPt v2, hard subset (Artificial Analysis run) | 24.5% | 71.5% | GPT-6.2 Astra |
| GPQA Diamond (Epoch AI run) | 90.2% | 96.8% | GPT-6.2 Astra |
| ARC-AGI | 91.0% | 98.5% | GPT-6.2 Astra |
| ARC-AGI-2 | 65.8% | 95.8% | GPT-6.2 Astra |
| SimpleBench | Not measured | 77.3% | Only GPT-6.2 Astra measured |
| Chess puzzles | 21.0% | 78.0% | GPT-6.2 Astra |
| Mystery game puzzles | 8.0% | 95.0% | GPT-6.2 Astra |
| EBR-bench | Not measured | 92.4% | Only GPT-6.2 Astra measured |
| LMArena Text: hard prompts | 1490 | 1425 | GLM 5.3 Flash |
| LMArena Text: expert | 1513 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: hard prompts in English | 1491 | 1426 | GLM 5.3 Flash |
| LMArena Vision: diagrams | 1320 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Vision: homework | 1325 | Not measured | Only GLM 5.3 Flash measured |
| Humanity's Last Exam (Scale run) | Not measured | 66.9% | Only GPT-6.2 Astra measured |
| GoBench | Not measured | 2493 | Only GPT-6.2 Astra measured |
| Neither model has a result on the other 4 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index03
| Benchmark | GLM 5.3 Flash | GPT-6.2 Astra | Leader |
|---|---|---|---|
| OTIS Mock AIME | 93.9% | 100.0% | GPT-6.2 Astra |
| FrontierMath, tiers 1 to 3 | 55.8% | 92.6% | GPT-6.2 Astra |
| FrontierMath, tier 4 | 17.1% | 100.0% | GPT-6.2 Astra |
| ProofBench | 21.0% | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: math | 1508 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: mathematical work | 1497 | Not measured | Only GLM 5.3 Flash measured |
| MathArena | Not measured | 92.9% | Only GPT-6.2 Astra measured |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index31
| Benchmark | GLM 5.3 Flash | GPT-6.2 Astra | Leader |
|---|---|---|---|
| MMLU Pro | 86.1% | Not measured | Only GLM 5.3 Flash measured |
| LegalBench (Vals AI) | 83.9% | Not measured | Only GLM 5.3 Flash measured |
| TaxEval (Vals AI) | 75.6% | Not measured | Only GLM 5.3 Flash measured |
| AA-Omniscience | 0.7 | 50.0 | GPT-6.2 Astra |
| SimpleQA Verified | Not measured | 78.5% | Only GPT-6.2 Astra measured |
| LMArena Text: science | 1488 | 1415 | GLM 5.3 Flash |
| LMArena Text: medicine and healthcare | 1476 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: legal and government | 1479 | 1415 | GLM 5.3 Flash |
| LMArena Text: business and finance | 1461 | 1413 | GLM 5.3 Flash |
| Neither model has a result on the other 12 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index06
| Benchmark | GLM 5.3 Flash | GPT-6.2 Astra | Leader |
|---|---|---|---|
| Design Arena: websites | 1252 | 1333 | GPT-6.2 Astra |
| Design Arena: UI components | 1289 | 1341 | GPT-6.2 Astra |
| Design Arena: game dev | 1253 | 1501 | GPT-6.2 Astra |
| Design Arena: data viz | 1254 | Not measured | Only GLM 5.3 Flash measured |
| Design Arena: 3D | 1281 | 1430 | GPT-6.2 Astra |
| Design Arena: all code categories | 1259 | 1389 | GPT-6.2 Astra |
| Design Arena: SVG | 1265 | Not measured | Only GLM 5.3 Flash measured |
| Design Arena: ASCII art | 1261 | Not measured | Only GLM 5.3 Flash measured |
| LiveBench design | 73.7% | 82.8% | GPT-6.2 Astra |
GeneralLike-for-likePublic boards, shown as published, kept out of the index102
| Benchmark | GLM 5.3 Flash | GPT-6.2 Astra | Leader |
|---|---|---|---|
| Vals Index | 50.9% | 67.7% | GPT-6.2 Astra |
| LiveBench data analysis | 76.4% | 84.1% | GPT-6.2 Astra |
| SAGE (Vals AI) | 89.3% | Not measured | Only GLM 5.3 Flash measured |
| MedScribe (Vals AI) | 88.9% | Not measured | Only GLM 5.3 Flash measured |
| Long context reasoning (AA-LCR) | 80.0% | Not measured | Only GLM 5.3 Flash measured |
| GDPval-AA | 1648 | Not measured | Only GLM 5.3 Flash measured |
| Surface Evolver Bench | 52.5% | Not measured | Only GLM 5.3 Flash measured |
| Furniture assembly | Not measured | 89.2% | Only GPT-6.2 Astra measured |
| LMArena Text | 1471 | 1414 | GLM 5.3 Flash |
| LMArena Text: instruction following | 1480 | 1423 | GLM 5.3 Flash |
| LMArena Text: creative writing | 1442 | 1417 | GLM 5.3 Flash |
| LMArena Vision | 1296 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: English | 1477 | 1417 | GLM 5.3 Flash |
| LMArena Text: non-English | 1461 | 1404 | GLM 5.3 Flash |
| LMArena Text: Chinese | 1527 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: German | 1457 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: Japanese | 1450 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: Korean | 1427 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: French | 1487 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: Spanish | 1478 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: Russian | 1469 | 1437 | GLM 5.3 Flash |
| LMArena Text: Polish | 1451 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Text: multi-turn | 1474 | 1425 | GLM 5.3 Flash |
| LMArena Text: longer queries | 1482 | 1431 | GLM 5.3 Flash |
| LMArena Text: writing and language | 1457 | 1440 | GLM 5.3 Flash |
| LMArena Text: entertainment and media | 1442 | 1413 | GLM 5.3 Flash |
| LMArena Vision: English | 1300 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Vision: Chinese | 1357 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Vision: reading text | 1306 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Vision: humor | 1291 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Vision: recognizing things | 1275 | Not measured | Only GLM 5.3 Flash measured |
| LMArena Vision: creative writing | 1272 | Not measured | Only GLM 5.3 Flash measured |
| MultiChallenge | Not measured | 73.2% | Only GPT-6.2 Astra measured |
| Neither model has a result on the other 16 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GLM 5.3 Flash | GPT-6.2 Astra | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
GLM 5.3 Flash costs less
Repository review
50K input + 3K output tokens
GLM 5.3 Flash costs less
Agent loop
200K input + 10K output tokens
GLM 5.3 Flash costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Z.ai GLM 5.3 Flash | OpenAI GPT-6.2 Astra |
|---|---|---|
| Maker | Z.ai | OpenAI |
| Weights | Open weights | No public weights listed |
| Released | August 26, 2026 | November 17, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $0.15 | $8.00 |
| Output $/Mtok | $0.50 | $40.00 |
| Coding index | 56.3 | 87.0 |
| Independent measurements behind it | 6 | 4 |
| Cost per solved task | $0.02 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is GLM 5.3 Flash better than GPT-6.2 Astra for coding?
Pick GPT-6.2 Astra for coding: it scores 87.0 on the Vibe Coding Index against 56.3 for GLM 5.3 Flash, though on 4 benchmarks against 6. Pick GLM 5.3 Flash for output tokens 80.0x cheaper ($0.50 vs $40.00 per million) and work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.
Which is cheaper, GLM 5.3 Flash or GPT-6.2 Astra?
GLM 5.3 Flash costs $0.15 per million input tokens and $0.50 per million output tokens; GPT-6.2 Astra costs $8.00 and $40.00. Cost per solved task is known for GLM 5.3 Flash only.
Which has the bigger context window, GLM 5.3 Flash or GPT-6.2 Astra?
GPT-6.2 Astra, with 1.05M tokens against 1.05M.
How many benchmarks is the GLM 5.3 Flash vs GPT-6.2 Astra comparison based on?
56 benchmarks have a result for both models as of November 22, 2026. GLM 5.3 Flash's coding index rests on 6 independent benchmarks and GPT-6.2 Astra's on 4. Where only one model has a result, the page says so instead of naming a winner.