Model comparison
GLM 5.2 vs GLM 5.3 Flash
Two Z.ai tiers compared on the figures that decide which one a workload actually needs.
Catalog record checked September 3, 2026Individual provider fields may changeEvidence confidence: High — see receipts below
Z.ai
GLM 5.2
Balanced · Open weights
Z.ai
GLM 5.3 Flash
Balanced · Open weights
| Specification | GLM 5.2 | GLM 5.3 Flash |
|---|---|---|
| Provider | Z.ai | Z.ai |
| Tier | Balanced | Balanced |
| Context window | 1M | Winner: 1.05M |
| Max output | 128K | Winner: 131K |
| Input / 1M tokens | $1.40 | Winner: $0.15 |
| Output / 1M tokens | $4.40 | Winner: $0.50 |
| Weights | Open | Open |
| Parameters | open MoE | 320B total / 18B active (MoE) |
| Reasoning levels | low, high, max | Not verifiedUnverified |
| Modalities | text | text, image, video |
| License | Not disclosedUnverified | MIT |
| API model id | glm-5.2 | Not publishedUnverified |
| Released | March 15, 2026 | August 26, 2026 |
| Artificial Analysis Intelligence Index (2026-08-14) | 50 | Winner: 57 |
| SWE-bench Verified (2026-09-01) | Not verifiedUnverified | 92 |
| Terminal-Bench 2.1 (2026-08) | Not verifiedUnverified | 84.3 |
| DeepSWE 1.1 (2026-08) | Not verifiedUnverified | 63.4 |
| Humanity's Last Exam (2026-08) | Not verifiedUnverified | 55.3 |
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
- 2026-09-01: vals.ai SWE-bench Verified leaderboard (independent, mini-swe-agent bash-only); not published by Z.aiReal GitHub issue resolution: does the model's patch pass the hidden tests. Comparability: comparable with caveat — Post-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.
- 2026-08-26: Artificial AnalysisComposite index blending reasoning, knowledge, and coding evals into one 0–100 score. Comparability: directly comparable — AA occasionally rebaselines the index scale between snapshots — a score captured on one date is only comparable to same-snapshot scores (check measuredAt).
- 2026-08: Z.ai GLM-5.3-Flash blog (vendor, mini-swe-agent, 400K context)Agentic software engineering: pass@1 on real repo tasks with reported cost per completed task. Comparability: directly comparable
- 2026-08: GLM-5.3-Flash HF model card (vendor, with tools, full set)Agentic terminal work: multi-step tasks executed in a sandboxed shell environment. Comparability: comparable with caveat — Scores across Terminal-Bench major versions (v1 vs v2) are NOT comparable; harness configuration (container, time limits) also shifts results.
Pricing tiers: GLM 5.2: Z.ai's per-token table lists $1.40/$4.40 per MTok (same list as 5.3). Open-weight, text-only; 1M context, 128K max output per Z.ai docs. · GLM 5.3 Flash: $0.15/$0.50 per MTok is the standard first-party and third-party rate (Z.ai docs; GMI, Novita, Together). A 50% promo tier runs $0.075/$0.25 — confirm which rate your account quotes.
GLM 5.2
GLM 5.2 is Zhipu's previous-generation open-weight model with strong cost-performance.
Best for
- Open-weight deployments
- Reasoning
- Cost-sensitive work
Watch out
Superseded by GLM 5.3 (Coding Plan 5.2/5.1 requests route to 5.3). Parameter count: sources disagree.
GLM 5.3 Flash
GLM 5.3 Flash is Z.ai's natively multimodal open-weight workhorse — 1M context, hybrid sparse/linear attention, near-flagship Intelligence Index at a fraction of the cost.
Best for
- Cost-efficient long-context
- Multimodal input
- Coding agents
Watch out
Self-hosting needs ~186GB GPU memory at 4-bit (multi-GPU). Third-party hosts charge more than Z.ai's own API.
When the cheaper one wins
GLM 5.3 Flash is cheaper on output at $0.50 per million tokens against $4.40 for GLM 5.2 — about 8.8×. 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 pickerEvidence 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.
- 4/5 core specs verified on both sides — Not published for at least one side: 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 8 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.
- GLM 5.2: HuggingFace — Z.ai (GLM) (accessed 2026-08-29)
- GLM 5.2: Z.ai (accessed 2026-08-29)
- GLM 5.3 Flash: Z.ai — GLM-5.3-Flash announcement (accessed 2026-08-29)
- GLM 5.3 Flash: getdeploying — GLM-5.3-Flash reference (accessed 2026-08-29)
- GLM 5.3 Flash: GMI Cloud — GLM-5.3-Flash analysis (accessed 2026-08-29)
Related comparisons
- Gemini 3.7 Flash vs GLM 5.3 Flash
- GLM 5.2 vs Grok 4.5
- Amazon Nova 2 Pro vs GLM 5.2
- Amazon Nova 2 Pro vs GLM 5.3 Flash
- Claude Sonnet 5 vs GLM 5.2
- Claude Sonnet 5 vs GLM 5.3 Flash
Diving deeper on one model? GLM 5.2 · GLM 5.3 Flash
Common questions
GLM 5.2 vs GLM 5.3 Flash
Answered from the verified figures on this page rather than general guidance.
Is GLM 5.2 or GLM 5.3 Flash cheaper for input?
GLM 5.3 Flash is cheaper at $0.15 per million input tokens, against $1.40 for GLM 5.2 — roughly 9.3× the price. Output tokens usually dominate a real bill, so weigh the output rate more heavily than the input rate. These are base rates; GLM 5.2 has tiered pricing: Z.ai's per-token table lists $1.40/$4.40 per MTok (same list as 5.3). Open-weight, text-only; 1M context, 128K max output per Z.ai docs. GLM 5.3 Flash has tiered pricing: $0.15/$0.50 per MTok is the standard first-party and third-party rate (Z.ai docs; GMI, Novita, Together). A 50% promo tier runs $0.075/$0.25 — confirm which rate your account quotes.
Is GLM 5.2 or GLM 5.3 Flash cheaper for output?
GLM 5.3 Flash is cheaper at $0.50 per million output tokens, against $4.40 for GLM 5.2 — roughly 8.8× the price. Output tokens usually dominate a real bill, so weigh the output rate more heavily than the input rate. These are base rates; GLM 5.2 has tiered pricing: Z.ai's per-token table lists $1.40/$4.40 per MTok (same list as 5.3). Open-weight, text-only; 1M context, 128K max output per Z.ai docs. GLM 5.3 Flash has tiered pricing: $0.15/$0.50 per MTok is the standard first-party and third-party rate (Z.ai docs; GMI, Novita, Together). A 50% promo tier runs $0.075/$0.25 — confirm which rate your account quotes.
Which has the larger context window, GLM 5.2 or GLM 5.3 Flash?
GLM 5.3 Flash accepts 1.05M tokens against 1M for GLM 5.2. This only matters if you routinely send very long documents or large codebases.
Should I use GLM 5.2 or GLM 5.3 Flash?
Both sit in the balanced tier, so the choice usually comes down to price and context rather than capability. GLM 5.2 suits open-weight deployments; GLM 5.3 Flash suits cost-efficient long-context.
Next step
Choosing between them
The verified figures that separate this pair, computed from the catalog rather than restated boilerplate.
- Input price: GLM 5.3 Flash costs $0.15 per 1M tokens versus $1.40 for GLM 5.2 — a 9.3x difference at the headline tier.
- Context: GLM 5.3 Flash takes 1.05M against 1M for GLM 5.2 — only decisive if your prompts approach the smaller window.
- Measured capability: GLM 5.3 Flash leads Artificial Analysis Intelligence Index 57 to 50 (measured 2026-08-14).
Choosing a harness rather than a model — Cursor, Copilot, Claude Code, Windsurf, or Lovable? Compare agentic harnesses · Latest releases.