Across 9 comparable benchmarks, GLM-5.2 has better reported results on 5 and worse on 4. 1 further shared result is listed below, each with the reason it is not counted.
Moderate evidence9 of 10 shared results countedHow we compare
Price basis: Standard rates vs international rates. The price section below shows the applicable periods.
| Fact | GPT-5.3 Codex | GLM-5.2 |
|---|---|---|
| Context window | 400K tokens | 1M tokens |
| Input / 1M | $1.75 | $1.40 |
| Output / 1M | $14.00 | $4.40 |
| Blended (4:1) | $3.07 | $1.18 |
| Status | Deprecated | Available |
| Released | February 5, 2026 | June 13, 2026 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 5 shared benchmarks grouped by capability on a 0–100 scale. How averages are built
GPT-5.3 Codex ahead on 3 of 5 counted
GLM-5.2 ahead on 1 of 1 counted
No fixed-scale average — see the rows in Benchmark results
No shared comparable results for Abstract reasoning, Mathematics, Computer & tool use, Research & search, Long context, Multimodal, Professional work & productivity, Cybersecurity.
Published scores, with matching modes where available. Gaps show GPT-5.3 Codex relative to GLM-5.2. Reading the results
| Benchmark | GPT-5.3 Codex | GLM-5.2 | Gap |
|---|---|---|---|
| APEX-SWERepository Engineering · Pass@1 | 41.5Thinking High · tools | 37.4≠Max · tools | +4.1 pts#29 vs #37 of 61 |
| SWE-Bench Pro - PublicRepository Engineering · Accuracy | 56.8Extra-High · no tools | 62.1≠Thinking · tools | −5.3 pts#27 vs #15 of 66 |
| Terminal Bench HardAgentic Development · Accuracy | 53.0Extra-High · tools | 50.8≠Max · tools | +2.2 pts#9 vs #12 of 157 |
| Vibe Code Bench v1.1Code Generation & Editing · Pass rate (%) | 61.8Extra-High · tools | 64.0≠Max · tools | −2.2 pts#58 vs #57 of 121 |
| WeirdML v2ML Engineering · Average accuracy across 17 tasks (%) | 79.3Normal · tools | 67.3≠Thinking High · tools | +12.0 pts#6 vs #15 of 39 |
| Vending-Bench 2Logical Planning · Money Balance (USD)Not counted · Speed and cost telemetry is reported, not scored | 5940Unspecified Reasoning · tools | 8314Unspecified Reasoning · tools | 0.71×#29 vs #19 of 64 |
| CritPtScientific Reasoning · Score | 16.9Extra-High · no tools | 20.9≠Max · no tools | 0.81×#36 vs #25 of 134 |
| ECICapability Indices · ECI score (capability index, higher is better) | 157Reported best | 152Reported best | 1.03×#18 vs #44 of 168 |
| LiveBenchCross-capability Suites · Accuracy | 72.8Thinking High · no tools | 73.2≠Normal · no tools | −0.4 pts#46 vs #44 of 106 |
| Text Arena (Coding)Preference Arenas · Arena Score | 1407Normal · no tools | 1593≠Max · no tools | 0.88×#24 vs #4 of 28 |
Repository Engineering · Pass@1
Repository Engineering · Accuracy
Agentic Development · Accuracy
Code Generation & Editing · Pass rate (%)
ML Engineering · Average accuracy across 17 tasks (%)
Logical Planning · Money Balance (USD)
Not counted · Speed and cost telemetry is reported, not scored
Scientific Reasoning · Score
Capability Indices · ECI score (capability index, higher is better)
Cross-capability Suites · Accuracy
Preference Arenas · Arena Score
A dash means no result is available.
Time-of-day providers use peak-hour rates; fixed-price providers use standard rates. Off-peak discounts remain in the published details.
Drag the handle above the bar for cache hits, the one below it for the input : output mix · 5% steps
Blended price = effective input rate at 90% cache hits × 80% + output rate × 20%
GPT-5.3 Codex costs160%morethan GLM-5.2 per million tokens — 2.6× the price.Using standard rates vs international rates at this token mix.
GPT-5.3 Codex: Standard pricing. OpenAI official pricing (developers.openai.com/api/docs/pricing) verified 2026-11-13; standard tier; Deprecated; shutdown 2027-04-01 (replacement: gpt-6-sol)
GLM-5.2: International pricing. Z.ai international price (docs.z.ai/guides/overview/pricing) verified 2026-10-22 via browser
| Price / 1M tokens | GPT-5.3 Codex | GLM-5.2 | Ratio |
|---|---|---|---|
| Input | $1.75 | $1.40 | 1.25× |
| Output | $14.00 | $4.40 | 3.18× |
| Cached input | $0.175 | $0.26 | 0.67× |
| Cache write | — | — | — |
| Blended | $3.07 | $1.18 | 2.60× |
Estimates use the selected token mix. Cache writes, tool fees and long-context surcharges are excluded.
Reported cost per task, paired with the score from the same run. Settings and sources may differ; costs do not affect the verdict.
| Benchmark | GPT-5.3 Codex | GLM-5.2 |
|---|---|---|
| Agent Arena | Not reported | $0.8461 / task Median · Net improvement index (percentage points vs. average model): 0.24 Max (With Tools + Internet) |
| Agent Arena Index | Not reported | $0.859 / task Median · Agent Arena Index rating (index points): 1090 Max (With Tools + Internet) |
| Agents' Last Exam v1.0 | Not reported | |
| AIME 2026 | Not reported | |
| AutomationBench | Not reported | |
| DeepSWE | Not reported |
| Attribute | GPT-5.3 Codex | GLM-5.2 |
|---|---|---|
| Availability | Deprecated | Available |
| Provider | OpenAI | Zhipu AI |
| Model type | Coding model | Reasoning model |
| First release | February 5, 2026 | June 13, 2026 |
| Context window | 400K tokens | 1M tokens |
| Max output | 128K tokens | 128K tokens |
| Total parameters | — | 753.3B |
| Active parameters | — | 40B |
| Architecture | Undisclosed | Mixture of Experts (MoE) |
| Documented capabilities | Reasoning, Multilingual, Coding | Reasoning, Multilingual, Coding |
| Input modalities | Text, Image | Text |
| Output modalities | Text | Text |
| Weights | Closed / API only | Open weights |
| Licence | — | MIT License |
| Commercial use | Proprietary | Commercial use allowed |
These are reported-score comparisons, not controlled head-to-head tests.
What people found comparing these models, or using either one. Comments also appear on the models' own pages.
Used GPT-5.3 Codex or GLM-5.2? Sign in with Google to share what worked, what did not, and which API or setup you tested.
No community comments on either model yet.
Alternatives with the most benchmark overlap for this matchup.