vs
0:5
overall capability wins (0 ties across 5 scored). DeepSeek V4.1 Flash leads overall.
- Reasoning0:1
- Coding0:1
- Pricing3:0
- Specs0:3
Verdict
DeepSeek V4.1 Flash leads coding (widest gap: +38.0 on SciCode); DeepSeek V4.1 Flash offers ~7.8x larger context (1M vs 128K); DeepSeek V4.1 Flash allows ~46.9x more max output (384K vs 8K).
- DeepSeek V4.1 Flash leads coding (widest gap: +38.0 on SciCode)
- DeepSeek V4.1 Flash offers ~7.8x larger context (1M vs 128K)
- DeepSeek V4.1 Flash allows ~46.9x more max output (384K vs 8K)
- DeepSeek Coder V2 Lite is ~4.2x cheaper on a blended token basis than DeepSeek V4.1 Flash
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- DeepSeek Coder V2 Lite
- $0.15
- DeepSeek V4.1 Flash
- $0.60
DeepSeek Coder V2 Lite is about 4.0x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
DeepSeek
DeepSeek
Family
DeepSeek
DeepSeek
License
DeepSeek
MIT
Open weights
Yes
Yes
Release
Jun 17, 2024
Sep 10, 2026
Knowledge cutoff
-
-
API / provider
DeepSeek / Together (ref.)
DeepSeek
Modalities
text → text
text, image → text
Specs
Context window
128K
1M
Max output
8K
384K
Parameters
16B (2.4B act.)
552B (16B act.)
Pricing
Input $/1M
$0.10
$0.30
Output $/1M
$0.20
$1.20
Blended $/1M (3∶1)
$0.13
$0.52
Speed
tok/s
110
-
TTFT (s)
0.28
-
Reasoning
MMLU-Pro
42.9
—
GPQA Diamond
31.9
—
Humanity's Last Exam
5.4
39.2
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
—
46.4
AA-LCR v1.1
—
84.0
CritPt
—
14.3
MMMU-Pro
—
77.0
IFBench
—
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
15.8
—
Terminal-Bench 2.1
—
—
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
36.8
—
SciCode
13.9
51.9
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
—
1,620
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
—
—
| Metric | DeepSeek Coder V2 Lite | DeepSeek V4.1 Flash | Delta |
|---|---|---|---|
| Identity | |||
| Organization | DeepSeek | DeepSeek | — |
| Family | DeepSeek | DeepSeek | — |
| License | DeepSeek | MIT | — |
| Open weights | Yes | Yes | — |
| Release | Jun 17, 2024 | Sep 10, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | DeepSeek / Together (ref.) | DeepSeek | — |
| Modalities | text → text | text, image → text | — |
| Specs | |||
| Context window | 128K | 1M | B +872K |
| Max output | 8K | 384K | B +376K |
| Parameters | 16B (2.4B act.) | 552B (16B act.) | B +536 |
| Pricing | |||
| Input $/1M | $0.10 | $0.30 | A +$0.20 |
| Output $/1M | $0.20 | $1.20 | A +$1 |
| Blended $/1M (3∶1) | $0.13 | $0.52 | A +$0.40 |
| Speed | |||
| tok/s | 110 | - | — |
| TTFT (s) | 0.28 | - | — |
| Reasoning | |||
| MMLU-Pro | 42.9 | — | — |
| GPQA Diamond | 31.9 | — | — |
| Humanity's Last Exam | 5.4 | 39.2 | B +33.8 pts |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | — | 46.4 | — |
| AA-LCR v1.1 | — | 84.0 | — |
| CritPt | — | 14.3 | — |
| MMMU-Pro | — | 77.0 | — |
| IFBench | — | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | 15.8 | — | — |
| Terminal-Bench 2.1 | — | — | — |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | 36.8 | — | — |
| SciCode | 13.9 | 51.9 | B +38.0 pts |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | — | 1,620 | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | — | — | — |
DeepSeek Coder V2 Lite
- Context
- 128K / 8K out
- Parameters
- 16B (2.4B act.)
- Price
- $0.10 / $0.20
- Speed
- 110 tok/s · 0.28s TTFT
- Modalities
- text → text
- License
- DeepSeek
16B MoE Lite sibling of DeepSeek Coder V2 — open coding at much lower GPU cost than the 236B MoE.
DeepSeek V4.1 Flash
- Context
- 1M / 384K out
- Parameters
- 552B (16B act.)
- Price
- $0.30 / $1.20 peak · $0.15 / $0.60 off-peak
- Speed
- —
- Modalities
- text, image → text
- License
- MIT
New default DeepSeek Flash tier (Sep 10, 2026), the smallest model in a new Causal Encoder-Decoder architecture family: 552B total parameters activating 8B (prefill) / 16B (decode) per token, MIT-licensed weights on Hugging Face. Replaces DeepSeek-V4-Flash and V4-Flash-Vision-Exp, which are retired but still routed to this model for compatibility. DeepSeek's own reported numbers (GPQA Diamond 90.9, Terminal-Bench 2.1 90.6, text-only HLE 39.1) are self-measured and not directly comparable to this catalog's Artificial-Analysis-sourced cells. API pricing: $0.30 in / $1.20 out peak, $0.15 in / $0.60 out off-peak (per 1M tokens); off-peak cache-hit input is $0.003.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.