Side-by-side comparison · models
GLM-4.7 vs GLM-5
What we have checked about each, line by line. Every value shows where it came from and when we last confirmed it.
Models
GLM-4.7
GLM vision model for visual reasoning, documents, and multimodal agents.
Models
GLM-5
General GLM flagship for coding, analysis, and tool-heavy engineering workflows.
Checked facts
Line by line
| Measure | GLM-4.7 | GLM-5 |
|---|---|---|
| Input price | $0.600 per million tokens | $1.00 per million tokens |
| Output price | $2.20 per million tokens | $3.20 per million tokens |
| Price on deepinfra | $0.064 per million cached input tokens | Not checked yet |
| Price on deepinfra | $0.320 per million input tokens | Not checked yet |
| Price on deepinfra | $1.40 per million output tokens | Not checked yet |
| Price on google-vertex | $0.600 per million input tokens | Not checked yet |
| Price on google-vertex | $2.20 per million output tokens | Not checked yet |
| Price on mancer | $0.350 per million cached input tokens | Not checked yet |
| Price on mancer | $0.700 per million input tokens | Not checked yet |
| Price on mancer | $2.50 per million output tokens | Not checked yet |
| Price on novita | $0.110 per million cached input tokens | $0.200 per million cached input tokens |
| Price on novita | $0.600 per million input tokens | $1.00 per million input tokens |
| Price on novita | $2.20 per million output tokens | $3.20 per million output tokens |
| Price on venice | $0.080 per million cached input tokens | $0.200 per million cached input tokens |
| Price on venice | $0.400 per million input tokens | $1.00 per million input tokens |
| Price on venice | $1.93 per million output tokens | $3.20 per million output tokens |
| Price on z-ai | $0.110 per million cached input tokens | $0.200 per million cached input tokens |
| Price on z-ai | $0.600 per million input tokens | $1.00 per million input tokens |
| Price on z-ai | $2.20 per million output tokens | $3.20 per million output tokens |
| Price on amazon-bedrock | Not checked yet | $1.00 per million input tokens |
| Price on amazon-bedrock | Not checked yet | $3.20 per million output tokens |
| Price on baidu | Not checked yet | $0.140 per million cached input tokens |
| Price on baidu | Not checked yet | $0.700 per million input tokens |
| Price on baidu | Not checked yet | $2.24 per million output tokens |
| Price on gmicloud | Not checked yet | $0.120 per million cached input tokens |
| Price on gmicloud | Not checked yet | $0.600 per million input tokens |
| Price on gmicloud | Not checked yet | $1.92 per million output tokens |
| Price on siliconflow | Not checked yet | $0.200 per million cached input tokens |
| Price on siliconflow | Not checked yet | $0.950 per million input tokens |
| Price on siliconflow | Not checked yet | $2.55 per million output tokens |
| Price on streamlake | Not checked yet | $0.120 per million cached input tokens |
| Price on streamlake | Not checked yet | $0.600 per million input tokens |
| Price on streamlake | Not checked yet | $1.92 per million output tokens |
| Latest version | z-ai/glm-4.7-20251222 | z-ai/glm-5-20260211 |
| Time to first token (ms) | 2,510 | 1,700 |
| Time to first token (ms) on deepinfra | 2,091 | Not checked yet |
| Time to first token (ms) on mancer | 1,060 | Not checked yet |
| Time to first token (ms) on novita | 4,540 | 2,116 |
| Time to first token (ms) on venice | 1,395 | 1,593 |
| Time to first token (ms) on z-ai | 3,922 | 1,453 |
| Output speed (tokens a second) | 39.5 | 65.5 |
| Output speed (tokens a second) on deepinfra | 36 | Not checked yet |
| Output speed (tokens a second) on mancer | 27.5 | Not checked yet |
| Output speed (tokens a second) on novita | 24 | 34 |
| Output speed (tokens a second) on venice | 30 | 54 |
| Output speed (tokens a second) on z-ai | 27 | 46 |
| Time to first token (ms) on amazon-bedrock | Not checked yet | 685 |
| Time to first token (ms) on baidu | Not checked yet | 1,845 |
| Time to first token (ms) on gmicloud | Not checked yet | 2,399 |
| Time to first token (ms) on siliconflow | Not checked yet | 2,316 |
| Time to first token (ms) on streamlake | Not checked yet | 1,762 |
| Output speed (tokens a second) on amazon-bedrock | Not checked yet | 115 |
| Output speed (tokens a second) on baidu | Not checked yet | 57 |
| Output speed (tokens a second) on gmicloud | Not checked yet | 47 |
| Output speed (tokens a second) on siliconflow | Not checked yet | 37 |
| Output speed (tokens a second) on streamlake | Not checked yet | 39 |
| Tokens served on OpenRouter in a day | 4.6B (for 10 Nov 2026) | 5.3B (for 10 Nov 2026) |
| Tool calls on OpenRouter in a day | 23.8K (for 10 Nov 2026) | 23.2K (for 10 Nov 2026) |
| Artificial Analysis Intelligence Index | 22.2 | 27.9 |
| Artificial Analysis Coding Index | 45.3 | Not checked yet |
| LMArena text rating (Elo) | 1,435 | 1,446 |
| LMArena web development rating (Elo) | 1,435 | 1,435 |
| Design Arena rating (Elo) | 1,222 | 1,244 |
| Gpqa diamond | 83.3% (for 22 Dec 2025) | 87.8% (for 11 Feb 2026) |
| Terminal bench | 33.4% (for 22 Dec 2025) | 52.4% (for 11 Feb 2026) |
| Simpleqa verified | 32.2% (for 22 Dec 2025) | Not checked yet |
| Chess puzzles | 21% (for 22 Dec 2025) | Not checked yet |
| Cl bench | 15.9% (for 22 Dec 2025) | 18.7% (for 11 Feb 2026) |
| Cl bench life | 10.9% (for 22 Dec 2025) | Not checked yet |
| Frontiermath 2025 02 28 private | 2.4% (for 22 Dec 2025) | 16.4% (for 11 Feb 2026) |
| Frontiermath tier 4 2025 07 01 private | 0% (for 22 Dec 2025) | 2.1% (for 11 Feb 2026) |
| Proofbench | 6% (for 22 Dec 2025) | Not checked yet |
| Simplebench | 47.7% (for 22 Dec 2025) | 53.2% (for 11 Feb 2026) |
| Otis mock aime 2024 2025 | 83.3% (saturated benchmark: the top models all score near the ceiling) (for 22 Dec 2025) | 80% (saturated benchmark: the top models all score near the ceiling) (for 11 Feb 2026) |
| Swe bench verified | Not checked yet | 72.1% (for 11 Feb 2026) |
| Arc agi 2 | Not checked yet | 4.9% (for 11 Feb 2026) |
| Arc agi | Not checked yet | 44.7% (for 11 Feb 2026) |
| Weirdml | Not checked yet | 48.2% (for 11 Feb 2026) |
A line appears only when at least one record has a value we confirmed from a source. Values we have not confirmed are left out, never guessed. Records of different types share fewer lines.
Add a third
Compare with one more record
How this works
Share it, or build your own
This page works for any two or three records in the directory, of any type. The link stays the same whichever order you name them in, so you can share it as it is. Change the records in the address, for example /directories/vs/claude-code-vs-cursor, or start from our editorial comparisons.