MiniMax-M3 vs Tencent Hy3 vs GPT-5.4 vs GPT-5.5
Pick two to four models to compare benchmark scores, prices, context windows and speed. The link updates as you pick, so you can share the exact comparison.
Cheapest blended
Tencent Hy3
$0.231 / 1M tokens
Largest context
GPT-5.4GPT-5.5
Tie at 1.05M tokens
Fastest output
MiniMax-M3GPT-5.4
Tie at 88 tokens/s
Top on BrowseComp
GPT-5.5
84.4%
Specifications
| Spec | MiniMax-M3MiniMax | Tencent Hy3Tencent | GPT-5.4OpenAI | GPT-5.5OpenAI |
|---|---|---|---|---|
| Provider | MiniMax | Tencent | OpenAI | OpenAI |
| Released | 1 Jun 2026 | Jul 2026 | 5 Mar 2026 | 24 Apr 2026 |
| Licence | Open weight | Open weight | Proprietary | Proprietary |
| Context window | 1M tokens | 262K tokens | 1.05M tokens | 1.05M tokens |
| Max output | 524K tokens | 128K tokens | 128K tokens | 128K tokens |
| Input / 1M | $0.30* | $0.132* | $2.50* | $5.00* |
| Output / 1M | $1.20 | $0.528 | $15.00 | $30.00 |
| Cached input / 1M | $0.06 | $0.033 | $0.25 | $0.50 |
| Speed | 88 tokens/s | 87 tokens/s | 88 tokens/s | 86 tokens/s |
| Input types | text, image, video | text | text, image | text, image |
| Best for | CodingAgentsLong contextMultimodal | CodingAgentsBudget | CodingAgents | CodingAgents |
* MiniMax-M3: Prices after a 'permanent 50% off' (list $0.60 / $2.40). Priority tier costs 1.5x standard.
* Tencent Hy3: Tencent Cloud TokenHub, Singapore region. OpenRouter now lists the same $0.132 / $0.528. An archived OpenRouter page from 8 Jul 2026 showed $0.20 / $0.80 plus a $0 endpoint; no free endpoint is listed now.
* GPT-5.4: There is no separate cache-write charge. Batch and Flex: $1.25 / $7.50. Fast mode: $5 / $30. Snapshot gpt-5.4-2026-03-05. Data residency +10%.
* GPT-5.5: There is no separate cache-write charge (pre-GPT-5.6 models). Batch and Flex: $2.50 / $15. Fast mode: 2.5x, $12.50 / $75. Snapshot gpt-5.5-2026-04-23. GPT-5.5 Pro (gpt-5.5-pro) costs $30 / $180. Data residency +10%.
Benchmark scores
Officially reported scores only. A model with no published score on a benchmark shows a line, not a zero.
Price per 1M tokens
Context window and speed
Notes
- MiniMax-M3
- MiniMax's latest pay-as-you-go language model: native multimodal (text, image, video), about 428B total and 23B active parameters, 1M context via MiniMax Sparse Attention. The MiniMax Community License grants rights for non-commercial use; commercial use requires showing "Built with MiniMax M3" and either a one-time notice to MiniMax or, above $20M yearly revenue, prior written authorization.
- Tencent Hy3
- Tencent Hy team's MoE model, 295B total and 21B active parameters (plus a 3.8B MTP layer), the full release after Hy3 Preview (late April 2026). Apache-2.0 with no extra conditions, so commercial use is unrestricted.
- GPT-5.4
- A frontier model for professional work and coding, and OpenAI's first general-purpose model with native computer use. It is now a lower-cost alternative to GPT-5.5.
- GPT-5.5
- A frontier model for coding and complex professional work, the flagship before GPT-5.6.
Still torn between two?
Test them on your own prompts.
Benchmarks are someone else’s work. On a free call we look at yours and agree how to check quality, cost and latency for the models you are weighing.