gpt-oss-20b vs Nex N2.5 Mini
Prices checked
gpt-oss-20b takes 4 of 5 rounds against Nex N2.5 Mini. It is also the cheaper of the two to run, at $0.02 per million tokens against $0.04.
OpenAIWins 4 roundsgpt-oss-20b
131k context · tool calling · open weights
$0.02/M tokensCheapest price, at TrustedRouter · 0.89¢ per 1,000 pages
Nex N2.5 Mini
262k context · reads images
$0.04/M tokensNex AGI's own price · 2.1¢ per 1,000 pages
Maker's price in (lower wins)
$0.02/M
$0.03/M
Maker's price out (lower wins)
$0.09/M
$0.10/M
Cheapest price (lower wins)
$0.02/M at TrustedRouter
none below the maker
Maker's price, blended (lower wins)
$0.04/M
$0.04/M
Cost per 1,000 pages (lower wins)
0.89¢
2.1¢
Context
131k tokens
262k tokens
Open weights
yes
no
Tool calling
yes
no
Structured output
yes
yes
Reads images
no
yes
Reasoning
yes
yes
Released
5 Aug 2025
8 Sept 2026
Providers selling it
47
4
Price we measure against
OpenRouter
the maker's own
Trains on your prompts (no is better)
not published
no provider to check
Providers selling both · 3
NanoGPT$0.04 / $0.02
Kilo Gateway$0.04 / $0.04
OpenRouter$0.04 / $0.04
gpt-oss-20b, then Nex N2.5 Mini, per million tokens.
Go further
Add a third or fourth model, or change the weighting, in the tool.
What moved this week
Impossibl · 3% dearer · 7.5¢ / $0.30 7¢ / $0.30
io.net Intelligence · 10% dearer · 5.3¢ / $0.17 4.5¢ / $0.16
Eden AI · 539% dearer · 7¢ / $0.25 0.9¢ / 4.5¢
Frequently asked
- Which is cheaper, gpt-oss-20b or Nex N2.5 Mini?
- gpt-oss-20b, at $0.02 per million tokens against $0.04 for Nex N2.5 Mini, blended 3 in : 1 out, read on 31 Oct 2026.
- Where is each one cheapest?
- gpt-oss-20b: $0.02 at TrustedRouter. Nex N2.5 Mini: nobody sells it below the maker's own price with a stated reason, so $0.04 is the number to beat.
- Can I buy both from one provider?
- Yes. 3 providers sell both: NanoGPT, Kilo Gateway, OpenRouter. The cheapest of them for the two together is NanoGPT.
- Which holds more context?
- Nex N2.5 Mini, at 262k tokens against 131k.
Prices exclude tax, read 31 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