Llama 4 Maverick vs Qwen2.5 7B Instruct
Prices checked
Qwen2.5 7B Instruct takes 3 of 5 rounds against Llama 4 Maverick. It is also the cheaper of the two to run, at $0.05 per million tokens against $0.34.
MetaLlama 4 Maverick
128k context · tool calling · reads images · open weights
$0.34/M tokensCheapest price, at Jiekou · $0.15 per 1,000 pages
Alibaba CloudWins 3 roundsQwen2.5 7B Instruct
131k context · tool calling · open weights
$0.05/M tokensCheapest price, at SiliconFlow · 3.3¢ per 1,000 pages
Maker's price in (lower wins)
$0.19/M
$0.17/M
Maker's price out (lower wins)
$0.65/M
$0.70/M
Cheapest price (lower wins)
$0.34/M at Jiekou
$0.05/M at SiliconFlow
Maker's price, blended (lower wins)
$0.30/M
$0.31/M
Cost per 1,000 pages (lower wins)
$0.15
3.3¢
Context
128k tokens
131k tokens
Open weights
yes
yes
Tool calling
yes
yes
Structured output
not published
not published
Reads images
yes
no
Reasoning
no
no
Released
5 Apr 2025
1 Sept 2024
Providers selling it
23
10
Price we measure against
OpenRouter
the maker's own
Trains on your prompts (no is better)
not published
not published
Providers selling both · 4
DeepInfra$0.35 / $0.06
Eden AI$0.35 / $0.06
Jiekou$0.34 / $0.07
Novita$0.41 / $0.07
Llama 4 Maverick, then Qwen2.5 7B Instruct, per million tokens.
Go further
Add a third or fourth model, or change the weighting, in the tool.
What moved this week
TrustedRouter · 0.6% cheaper · $0.28 / $0.87 $0.28 / $0.88
Featherless · 30% cheaper · $0.40 / $0.80 $0.68 / $0.80
Frequently asked
- Which is cheaper, Llama 4 Maverick or Qwen2.5 7B Instruct?
- Qwen2.5 7B Instruct, at $0.05 per million tokens against $0.34 for Llama 4 Maverick, blended 3 in : 1 out, read on 21 Oct 2026.
- Where is each one cheapest?
- Llama 4 Maverick: $0.34 at Jiekou. Qwen2.5 7B Instruct: $0.05 at SiliconFlow.
- Can I buy both from one provider?
- Yes. 4 providers sell both: DeepInfra, Eden AI, Jiekou, Novita. The cheapest of them for the two together is DeepInfra.
- Which holds more context?
- Qwen2.5 7B Instruct, at 131k tokens against 128k.
Prices exclude tax, read 21 Oct 2026 from each provider's own catalog. They are per M tokens, blended 3 in : 1 out, and a page is 600 tokens in and 60 out. Bars are scaled to the larger value in each row.
Related comparisons