MiniMax-M3 vs DeepSeek V4 Flash
Side-by-side comparison of API pricing, specs, benchmarks, and capabilities
MiniMax-M3 is 20.5% cheaper on blended cost ($0.525 vs $0.660/Mtok)
| Specification |
by MiniMax
3 providers:
Fireworks $0.300/$1.20
MiniMax $0.300/$1.20
Together $0.300/$1.20
|
by DeepSeek
3 providers:
DeepSeek $0.440/$1.32
Fireworks $0.140/$0.280
DeepInfra $0.090/$0.180
|
|---|---|---|
| Overview | ||
| Status | Current budget Open weights | Current budget |
| Released | Jun 1, 2026 | Apr 24, 2026 |
| Pricing per million tokens | ||
| Input | $0.300/Mtok | $0.440/Mtok |
| Output | $1.20/Mtok | $1.32/Mtok |
| Blended avg | $0.525/Mtok | $0.660/Mtok |
| Cached input | $0.060/Mtok | $0.014/Mtok |
| Price basis | — | DeepSeek bills this model at TWO rates depending on the hour. The tracked figure is the PEAK rate: $0.44 input / $0.014 cached input / $1.32 output per 1M tokens. Off-peak is exactly half: $0.22 / $0.007 / $0.66. Peak hours are 01:00-04:00 and 06:00-10:00 UTC; every other hour is off-peak. Peak is the headline here because DeepSeek defines off-peak as half of peak rather than the other way round, so peak is the list rate, and because a caller who does not schedule around the clock needs the published number to be a ceiling rather than a floor. Halve it for a workload that runs entirely outside those seven hours. Until 16:00 UTC on 2026-08-16 this model billed a single flat $0.14 / $0.0028 / $0.28. |
| Specifications | ||
| Context window | 1M tokens | 1M tokens |
| Parameters | Proprietary | Proprietary |
| Speed (TPS) | — | — |
| Modalities | ||
| Input |
textimagevideo
|
text
|
| Benchmarks source: Vellum LLM Leaderboard | ||
| Avg benchmark score | 63.1 | 77.5 |
| Perf / dollar | 120.2 | 117.4 |
| GPQA Diamond | 93 | 88.1 |
| SWE-Bench Verified | 45 | 79 vendor |
| MCP Atlas | 74.2 | 69 |
| BrowseComp | 83.5 | 85.9 |
| Humanity's Last Exam | 22 | 51.6 |
| ARC-AGI 2 | 20 | 18 vendor |
| AIME 2025 | 55 | 58 vendor |
| MMMLU | 76 | 75 vendor |
| BFCL | 65 | 62 vendor |
| HumanEval | 80.5 | 79 |
| MATH 500 | 70 | 68 vendor |
| Terminal-Bench 2.1 | 66 | — |
| LiveCodeBench | — | 91.6 |
| OSWorld-Verified | 70.1 | — |
| Independent composite scores each on its own scale — not part of the average above | ||
| AA Intelligence Index | 45.4 | 51.8 |
| AA Agentic Index | 36.1 | 48.4 |
| AA-Omniscience Index (−100–100) | 1.4 | -14.3 |
| GDPval-AA v2 | 1380.2 ELO | 1558.4 ELO |
| LMSYS Chatbot Arena | 1442 ELO | 1435.7 ELO |
| Providers | ||
| Available from |
Fireworks — $0.300/$1.20/Mtok
MiniMax — $0.300/$1.20/Mtok
Together — $0.300/$1.20/Mtok
|
DeepSeek — $0.440/$1.32/Mtok
Fireworks — $0.140/$0.280/Mtok
DeepInfra — $0.090/$0.180/Mtok
|
Cost at scale
1M tokens · 50/50 input/output| Volume | MiniMax-M3 | DeepSeek V4 Flash | Savings |
|---|---|---|---|
| 1M tokens | $0.52 | $0.66 | $0.14 (21.2%) |
| 10M tokens | $5.25 | $6.6 | $1.35 (20.5%) |
| 100M tokens | $52.5 | $66 | $13.5 (20.5%) |
| 1000M tokens | $525 | $660 | $135 (20.5%) |
When to pick which
distilled from the pricing and spec data above
Pick MiniMax-M3 if…
- cost dominates: $0.525/Mtok blended vs $0.660 — 20.5% less on the same 50/50 token mix
- you want open weights — self-host it, fine-tune it, or exit the API entirely; DeepSeek V4 Flash is closed
Pick DeepSeek V4 Flash if…
- your workload re-reads context (agents, RAG, long chats): cached input costs $0.014/Mtok — 97% off its list input price
- raw capability matters more than price: 77.5 vs 63.1 average benchmark score
Price movement
recorded changes since we began tracking each model · last checked Aug 28, 2026| Model | Changes | Latest move | Old → new $/M |
|---|---|---|---|
| MiniMax-M3 | 1 | price cut on Jun 17, 2026 | $0.6/$2.4 → $0.3/$1.2/Mtok |
| DeepSeek V4 Flash | 1 | price rise on Aug 16, 2026 | $0.14/$0.28 → $0.44/$1.32/Mtok |
Both models have re-priced since we began tracking — pricing here is active, so re-check before committing to a long-term choice.
See all recent price moves →
MODELPRICEWATCH.COM · 2026-08-28
Try MiniMax-M3 on MiniMax
The cheaper option here — MiniMax-M3 costs $0.525/Mtok blended on MiniMax.
Summary
MiniMax-M3 by MiniMax costs $0.300/Mtok input and $1.20/Mtok output, with a 1M-token context window. It supports text, image, video input and is available from 3 providers.
DeepSeek V4 Flash by DeepSeek costs $0.440/Mtok input and $1.32/Mtok output, with a 1M-token context window. It supports text input and is available from 3 providers.
On a blended cost basis, MiniMax-M3 is 20.5% cheaper than DeepSeek V4 Flash.
On benchmarks, DeepSeek V4 Flash scores higher (77.5 vs 63.1) on average. In terms of value, MiniMax-M3 has better performance per dollar (120.2 vs 117.4).
Note: Pricing is per million tokens. Actual costs vary with usage patterns, prompt caching, and batch discounts. Always verify against official provider pricing pages.
More comparisons
pairs sharing a model with this page
Solar Pro 4 vs DeepSeek V4 Flash
92% price gap
MiniMax-M3 vs GLM-5.2
76% price gap
MiniMax-M3 vs DeepSeek V4 Pro
74% price gap
MiniMax-M3 vs Gemini 2.5 Flash
38% price gap
DeepSeek V4 Flash vs Gemini 2.5 Flash
22% price gap
MiniMax-M3 vs Gemini 3.1 Flash-Lite
7% price gap
GLM-4.7-Flash vs Ministral 3 3B
100% price gap
Solar Pro 4 vs Gemini 3.5 Flash-Lite
94% price gap