DeepSeek V4.1 Flash 1017 vs GPT-6.1 Terra
DeepSeek V4.1 Flash 1017 is from DeepSeek. GPT-6.1 Terra is from OpenAI. DeepSeek V4.1 Flash 1017 costs $0.600 per million output tokens. GPT-6.1 Terra costs $5.00. GPT-6.1 Terra leads on the Epoch Capabilities Index: DeepSeek V4.1 Flash 1017 scores 158.8 and GPT-6.1 Terra scores 168.1. A typical request, about 1,000 words in and 400 words out, costs $0.0005 on DeepSeek V4.1 Flash 1017 and $0.0038 on GPT-6.1 Terra.
Prices updated . Benchmarks updated .
| DeepSeek V4.1 Flash 1017 | GPT-6.1 Terra | |
|---|---|---|
| Lab | DeepSeek | OpenAI |
| You send / 1M | $0.150 | $1.00 |
| It replies / 1M | $0.600 | $5.00 |
| Context | 1.05M | 1.05M |
| Launched | 2026-10-17 | 2026-10-26 |
| Weights | Open weights | Closed |
| Typical request | $0.0005 | $0.0038 |
| ECI | 158.8 | 168.1 |
| GPQA | 93.2% | 94.3% |
| ARC-AGI-2 | 78.1% | n/a |
| HLE | n/a | n/a |
| AIME | 99.4% | 100.0% |
| SimpleQA | 42.5% | 57.3% |
| CritPt | 24.0% | 28.0% |
| Arena | n/a | n/a |
| Arena web | n/a | 1,673 |
Which is cheaper?
DeepSeek V4.1 Flash 1017 costs less to run: $0.600 per million output tokens, against $5.00 for GPT-6.1 Terra. Input tokens are $0.150 per million for DeepSeek V4.1 Flash 1017 and $1.00 per million for GPT-6.1 Terra. A typical request, about 1,000 words in and 400 words out, costs $0.0005 on DeepSeek V4.1 Flash 1017 and $0.0038 on GPT-6.1 Terra.
Which scores higher?
GPT-6.1 Terra scores higher on Epoch Capabilities Index: 168.1 against 158.8 for DeepSeek V4.1 Flash 1017. That index is an independent composite from Epoch AI. Lab-reported scores are kept off this comparison. Context windows are 1.05M and 1.05M. Prices are US dollars per million tokens at current public rates.
Verdict
GPT-6.1 Terra scores higher on Epoch Capabilities Index (168.1 vs 158.8). DeepSeek V4.1 Flash 1017 costs less at $0.600 per million output tokens. Context is 1.05M for DeepSeek V4.1 Flash 1017 and 1.05M for GPT-6.1 Terra. Use the model pages for every score on file, including lab-reported numbers.
FAQ
Which is cheaper, DeepSeek V4.1 Flash 1017 vs GPT-6.1 Terra?
DeepSeek V4.1 Flash 1017 costs less to run: $0.600 per million output tokens, against $5.00 for GPT-6.1 Terra. Input tokens are $0.150 per million for DeepSeek V4.1 Flash 1017 and $1.00 per million for GPT-6.1 Terra. A typical request, about 1,000 words in and 400 words out, costs $0.0005 on DeepSeek V4.1 Flash 1017 and $0.0038 on GPT-6.1 Terra.
Which scores higher, DeepSeek V4.1 Flash 1017 vs GPT-6.1 Terra?
GPT-6.1 Terra scores higher on Epoch Capabilities Index: 168.1 against 158.8 for DeepSeek V4.1 Flash 1017. That index is an independent composite from Epoch AI. Lab-reported scores are kept off this comparison. Context windows are 1.05M and 1.05M. Prices are US dollars per million tokens at current public rates.
What is the verdict for DeepSeek V4.1 Flash 1017 vs GPT-6.1 Terra?
GPT-6.1 Terra scores higher on Epoch Capabilities Index (168.1 vs 158.8). DeepSeek V4.1 Flash 1017 costs less at $0.600 per million output tokens. Context is 1.05M for DeepSeek V4.1 Flash 1017 and 1.05M for GPT-6.1 Terra. Use the model pages for every score on file, including lab-reported numbers.
Open this pair in the comparison view
Sources and licenses
- OpenRouter: Prices, context windows, modalities and the model list come from the OpenRouter public API. Data: OpenRouter.
- Vercel AI Gateway: The no-training flag comes from the Vercel AI Gateway model list.
- models.dev: Lab launch dates, open-weights flags and knowledge cutoffs come from models.dev. Copyright (c) 2025 models.dev, MIT License.
- Epoch AI: Independent benchmark scores and the Epoch Capabilities Index (ECI) come from Epoch AI's Benchmarking Hub, used under CC BY 4.0. Epoch AI, Capabilities & benchmarking, epoch.ai/benchmarks. Scores Epoch marks as external keep their original license, and the source is named on each score.
- LMArena: Arena Elo (text, webdev and vision) comes from the LMArena leaderboard dataset (lmarena-ai/leaderboard-dataset), CC BY 4.0, leaderboard date 2026-10-28.
- Lab-reported scores, when an editor has added them, are the lab's own numbers. Each one links to the source and is labeled lab-reported. They are not used to rank the lists.
- What a score measures: the benchmark guide.