deepseek-flash vs gpt-6-astra
Parameters, independent leaderboard coverage, real-task reports and open-weight availability are shown side by side; missing coverage is not a zero.
Parameters
| Comparison item | deepseek-flash | gpt-6-astra |
|---|---|---|
| Model Type | LLM | LLM |
| Context window | 1,000,000 tokens | 1,050,000 tokens |
| Maximum output | 384,000 tokens | 128,000 tokens |
| Input/output modalities | text, image → text | text, image → text |
| Endpoint | openai · openai-response · anthropic | openai · openai-response |
| Open weights | Available | Unavailable |
| License | MIT | Proprietary hosted model |
Leaderboard coverage
Every independent board covering either model is retained; an em dash means that board has not published a figure for that model.
A row marked as retrieved on different dates holds two readings taken on different days; a board re-fits or changes version as a whole between readings, so those two figures are not directly comparable.
| Evaluation | deepseek-flash | gpt-6-astra |
|---|---|---|
| LMArena Text Arena — Arena score (Elo) | — | 1480 ±12 |
| LiveBench — Overall | 81.1 | 82.2 |
| Artificial Analysis Intelligence Index — Intelligence Index | 39 | 53 |
| SuperCLUE 智能指数 — 总分 (overall) | 71.81 | — |
| Epoch Capabilities Index (ECI) — General ECI | 155 (149 - 158) | 167 (163 - 172) |
| Vals Index — Accuracy (GDP-weighted finance, coding and legal) | 57.86 ±1.16 | 66.61 ±1.09 |
| LMArena Text Arena · Coding — Arena score (Elo) | — | 1543 ±23 |
| LiveBench · Coding — Coding | 80.0 | 80.4 |
| LiveBench · Agentic Coding — Agentic Coding | 77.3 | 57.3 |
| Artificial Analysis · Output Speed — Median output tokens/s | 220 | 58 |
| Artificial Analysis · GDPval-AA v2.1 — Agentic Real-World Work Tasks, (Elo-500)/2000 (%) | 55 | 52 |
| Artificial Analysis · Terminal-Bench v4.0 — Agentic Coding & Terminal Use (%) | 27 | 59 |
| Artificial Analysis · MLCR-AA — Medical Long Context Reasoning (%) Retrieved on different dates | 22.8 (retrieved 2026-09-22) | 35.0 (retrieved 2026-09-07) |
Real-task reports
deepseek-flash
- RemakeBench Launch 005: four Flash models on 14 fixtures — Fourteen fixtures spanning game building, 3D scenes, CAD, drawing and robot control, run on Qwen 3.8 Flash, DeepSeek V4.1 Flash, GLM 5.3 Flash and Gemini 3.8 Flash. Finding: No model won overall. DeepSeek V4.1 Flash delivered the best submitted Turbofan CAD result (three valid native solids, with a wrong-sign coupling and LP/HP interference) and did not pass the single-turn Rubik's Cube task. Limitation: Pass/fail labels are video judgments without a machine-readable validator, and single attempts do not establish reliability; the lab publishes no aggregate score or overall winner. Several DeepSeek clips reuse earlier records that keep V4 labels and the Vision exp setting; the two results quoted here are campaign runs added for this comparison.
- MiMo-V2.6-Flash, DeepSeek V4.1 Flash and Grok 4.7 on one game prompt — Build a playable browser game from a single prompt with Command Code's /design command. Finding: DeepSeek V4.1 Flash scored 9/10 for $0.0089 with a playable one-shot game. MiMo-V2.6-Flash also scored 9/10 at $0.005; Grok 4.7 scored 8/10 at $0.20 and needed iterations. Limitation: One prompt rated by the author, whose profile lists work for Command Code, the tool that ran the test; there is no rubric and no repeated run.
- DeepSeek V4.1 Flash and GLM 5.3 Flash recreating two game menus — Recreate the Psychonauts 2 and Metroid Prime main menus as closely as possible to the originals. Finding: DeepSeek V4.1 Flash used 412k and 466k tokens, three to four times GLM 5.3 Flash's 134k and 111k, and the author judged GLM's menus clearly better on both games. Limitation: Which menu is closer to the original is the author's own judgement on two prompts, without a scoring rubric.
- Switching a coding project from GPT-6 Astra to DeepSeek V4.1 Flash — Review an existing project, list its issues and fix them. Finding: DeepSeek V4.1 Flash found several issues, asked which to tackle first and began fixing them; Kimi's review still found problems, which were passed back to DeepSeek to correct. Limitation: One uncontrolled self-report. The author ran DeepSeek through WorkBuddy's free access, so the remark about spending nothing extra does not reflect API pricing.
- Step 5 Preview and DeepSeek V4.1 Flash on a 3D arcade game — Build Katamari Tiny, a 3D arcade game, from a single prompt. Finding: DeepSeek V4.1 Flash used 26,898 tokens against Step 5 Preview's 39,830 and, in the author's view, was faster and produced the better game. Limitation: One prompt judged by an author who names DeepSeek V4.1 Flash as their default model; the post gives no timings or quality rubric.
- DeepSeek V4.1 Flash and GLM-5.3-Flash on a Blender geometry task, run locally — Write a Blender script that builds a dodecahedron trapped inside a dodecahedron. Finding: Both models produced a correct mesh with zero intersections. DeepSeek V4.1 Flash took 20 min 15 s and 62,928 tokens at 51.8 tok/s, about a fifth of GLM-5.3-Flash's time and half its tokens, and its computed clearance matched the measured mesh (0.238); the author found GLM's picture nicer. Limitation: One local run per model on different hardware and precision: GLM ran as a third-party EXL3 quantisation on two DGX Sparks and DeepSeek as FP8 on four, so speed and token counts describe those builds, not the official weights or a hosted API.
- Three Flash models build the same koi pond locally — A koi pond build, run as a local head-to-head. Finding: All three finished cleanly with zero fixes. DeepSeek V4.1 Flash was fastest at 21 minutes and 53.3K tokens, a third of GLM-5.3-Flash's time, and the author ranked it first. Limitation: One local run per model; the post states neither the prompt nor the quantisation or runtime used.
- Step 5 Preview and DeepSeek V4.1 Flash on one Three.js scene — A Three.js scene of giant red mesas in a desert, with long shadows and a huge empty sky. Finding: DeepSeek V4.1 Flash's scene was more restrained but had clearly better ground and mesa textures and a smoother camera; Step 5 Preview added more atmosphere, a tracked sun, particle and dust effects and on-screen text. Limitation: One prompt judged by eye by the author; the post does not say how either model was accessed.
gpt-6-astra
- Cortex AI: GPT-6 Astra, Gemini 3.8 Flash and Fable 5.1 on two real robot tasks — Load a sandwich into a microwave and close the door; attach a 5 mm tip to a pipette. Finding: GPT-6 Astra scored highest on both tasks, 80% on the microwave and 70% on the pipette, against 13.3% and 30% for Gemini 3.8 Flash and 6.7% and 20% for Fable 5.1; the lab credits it with retrying until the pipette tip docked and with checking that the microwave door latched. Limitation: The post does not say how many attempts each model had; scores are partial-credit judgements by the lab's evaluator, and the authors attribute the gap mainly to weaker depth perception in the lower-scoring models.
- GPT-6 Astra, Sol and Luna on one SVG animation prompt — Generate an SVG animation of a pelican riding a bicycle, shown in H5. Finding: GPT-6 Astra took 6 min 3 s and used 16% of the five-hour quota, against 4 min 42 s and 2% for GPT-6 Sol; the author recommends Sol when quota matters. Limitation: One prompt; the quota shares are a subscription meter, not API cost, and the post scores no output quality for Astra.
- Switching a coding project from GPT-6 Astra to DeepSeek V4.1 Flash — The author's own coding project, first on GPT-6 Astra and then on DeepSeek V4.1 Flash. Finding: About a dozen GPT-6 Astra turns cost the author $130 in a day without meeting the requirements; they switched the project back to DeepSeek V4.1 Flash. Limitation: One uncontrolled self-report; the post does not describe what the Astra turns were asked to do.
- Four models build the Eiffel Tower in Three.js — Build the Eiffel Tower in Three.js. Finding: GPT-6 Astra cost $7.45 and took 8 minutes; the author preferred Claude Opus 5.5's result ($8.95, 10 minutes) for this task. Limitation: One task judged by eye by the author; the post does not say how the models were run or how the costs were counted.