GLM-4.6 (Reasoning) vs Gemini 3 Flash Preview (Reasoning)
Comparing 2 AI models · 5 benchmarks · Z AI, Google
Most Affordable
Go
Gemini 3 Flash Preview (Reasoning)
$0.50/1M
Highest Intelligence
Go
Gemini 3 Flash Preview (Reasoning)
89.8% GPQA
Best for Coding
Go
Gemini 3 Flash Preview (Reasoning)
41.0 Coding Index
Price Difference
1.1x
input cost range
Composite Indices
Intelligence, Coding, Math
Standard Benchmarks
Academic and industry benchmarks
Benchmark Winners
5 tests
Z
GLM-4.6 (Reasoning)
0
No clear wins
Go
Gemini 3 Flash Preview (Reasoning)
5
- GPQA
- MMLU Pro
- HLE
- LiveCodeBench
- AIME 2025
| Metric | Z GLM-4.6 (Reasoning) | Go Gemini 3 Flash Preview (Reasoning) |
|---|---|---|
| Pricing Per 1M tokens | ||
| Input Cost | $0.55/1M | $0.50/1M |
| Output Cost | $2.20/1M | $3.00/1M |
| Blended Cost 3:1 input/output ratio | $0.96/1M | $1.13/1M |
| Specifications | ||
| Organization Model creator | Z AI | |
| Release Date Launch date | Sep 30, 2025 | Dec 17, 2025 |
| Performance & Speed | ||
| Throughput Output speed | 133.6 tok/s | 217.7 tok/s |
| Time to First Token (TTFT) Initial response delay | 502ms | 11638ms |
| Latency Time to first answer token | 15477ms | 11638ms |
| Composite Indices | ||
| Intelligence Index Overall reasoning capability | 32.2 | 45.9 |
| Coding Index Programming ability | 28.4 | 41.0 |
| Math Index Mathematical reasoning | 86.0 | 97.0 |
| Standard Benchmarks | ||
| GPQA Graduate-level reasoning | 78.0% | 89.8% |
| MMLU Pro Advanced knowledge | 82.9% | 89.0% |
| HLE Hard language evaluation | 13.3% | 34.7% |
| LiveCodeBench Real-world coding tasks | 69.5% | 90.8% |
| MATH 500 Mathematical problems | — | — |
| AIME 2025 Advanced math competition | 86.0% | 97.0% |
| AIME (Original) Math olympiad problems | — | — |
| SciCode Scientific code generation | 38.4% | 50.6% |
| LCR Code review capability | 54.3% | 66.3% |
| IFBench Instruction-following | 43.4% | 78.0% |
| TAU-bench v2 Tool use & agentic tasks | 70.5% | 80.4% |
| TerminalBench Hard CLI command generation | ||