Compare / DeepSeek V3.1 Terminus vs Step 3.5 Flash
DeepSeek V3.1 Terminus vs Step 3.5 Flash
Which is cheaper, which is more capable, and which does better at each kind of work, from 15 results on 3 sources that measured both models.
DeepSeek V3.1 Terminus or Step 3.5 Flash? The short answer
- Step 3.5 Flash is cheaper: $0.15 against $0.45 per million tokens, 3.0 times less.
- Step 3.5 Flash scores higher on the Artificial Analysis Intelligence Index: 16.6 against 14.8.
- On our business benchmarks, DeepSeek V3.1 Terminus ranks higher for writing.
- On our business benchmarks, Step 3.5 Flash ranks higher for reasoning and knowledge.
At a glance
| DeepSeek V3.1 Terminus | Step 3.5 Flash | |
|---|---|---|
| Price per million tokens | $0.27 in · $1.00 out | $0.10 in · $0.30 out |
| Blended (3 in : 1 out) | $0.45 | $0.15 |
| Cheapest host, blended | $0.45 (SiliconFlow, FP8) | – |
| Intelligence Index | 14.8 | 16.6 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | – | – |
| Context window | 163,840 tokens | 262,144 tokens |
| Weights | Open | Open |
| Developer based in | China | China |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among models available today, for each kind of work, from our own benchmarks and licensed sources. The better rank is in green.
| Use case | DeepSeek V3.1 Terminus | Step 3.5 Flash |
|---|---|---|
| Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls | 84 of 161 | not measured |
| Reasoning and knowledgeHard questions across science, maths and general knowledge | 105 of 195 | 96 of 195 |
| WritingWhat people prefer in blind comparisons, and judged writing quality | 79 of 187 | 99 of 187 |
| Sticking to the factsSummarising without inventing things, and factual answers | not measured | 123 of 153 |
Every shared result
| Benchmark | Metric | DeepSeek V3.1 Terminus | Step 3.5 Flash |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,423 | 1,390 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,405 | 1,345 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,420 | 1,384 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,417 | 1,393 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,411 | 1,366 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 14.8 | 16.6 |
| Artificial Analysisreported | GPQA Diamond% of questions | 79.2% | 83.1% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 16.4% | 21.1% |
| Artificial Analysisreported | IFBench% of instructions | 57.0% | 64.6% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 69.3% | 50.0% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 31.8% | 27.3% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 37.1% | 94.4% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 32.0% | 42.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.37 | 0.37 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 54.0 | 45.9 |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.
Quick answers
Is DeepSeek V3.1 Terminus better than Step 3.5 Flash?
It depends on the task. Among models available today, on our benchmarks and licensed sources, DeepSeek V3.1 Terminus ranks higher for writing; Step 3.5 Flash ranks higher for reasoning and knowledge.
Which is cheaper, DeepSeek V3.1 Terminus or Step 3.5 Flash?
Step 3.5 Flash costs $0.15 per million tokens (three input to one output), against $0.45 for DeepSeek V3.1 Terminus.
Run DeepSeek V3.1 Terminus and Step 3.5 Flash on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.