opendunes

gpt-oss-120b

by OpenAIReleased 5 August 2025

gpt-oss-120b reads text and writes text, keeping about 130 pages in mind at once. OpenAI released it so you can download it and run it yourself, and it is small enough for a single high-end graphics card. You choose how long it thinks (low, medium or high), you can read its full working, and it can use the tools and apps you connect, browse the web and fill in the exact form you give it.

TakesText
GivesText
Open in chatCreate API keyCompare

What it costs

You pay for what each request uses. No subscription, no minimum.

≈ 0,017 DA per message3 000 DA buys about 170 000 messages

One message, the ordinary use

1 000 tokens sent× 5,99 DA per 1M0,0060 DA
500 tokens back× 22,63 DA per 1M0,011 DA
Total≈ 0,017 DA

Every rate

Meter↑ you send↓ comes back
Text tokens6,32 DA23,30 DA
Cached tokens read5,66 DA

per 1M tokens

Answer a quick question≈ 0,0066 DA
Reply to a customer email≈ 0,013 DA
Write a one-page draft≈ 0,022 DA
Translate a full page of text≈ 0,022 DA
Explain a stack trace and suggest a fix≈ 0,023 DA
When you pay
After the answer, from your balance. Nothing is held.
Why it varies
A longer question or a longer answer costs more.
After each one
The exact charge shows on the request.
DepositDeposits from 50 DA, with Edahabia or CIB.

Where your request runs

22 companies run this model. OpenDunes sends each request to one of them, picked by price and speed. If it fails, the next one takes it. DA per 1M tokens: ↑ what you send · ↓ what comes back.

PrimeIntellect…
↑ Sent, per 1M
6,32 DA
↓ Back, per 1M
23,30 DA
Cache read
5,66 DA
Latency
…
Throughput
…
Near AI…
↑ Sent, per 1M
9,99 DA
↓ Back, per 1M
49,93 DA
Cache read
4,99 DA
Latency
…
Throughput
…
VeniceUnited States…
↑ Sent, per 1M
9,99 DA
↓ Back, per 1M
49,93 DA
Cache read
9,99 DA
Latency
…
Throughput
…
CoreWeaveUnited States…
↑ Sent, per 1M
9,99 DA
↓ Back, per 1M
56,58 DA
Cache read
9,99 DA
Latency
…
Throughput
…
DekaLLMAsia…
↑ Sent, per 1M
9,99 DA
↓ Back, per 1M
59,91 DA
Cache read
9,99 DA
Latency
…
Throughput
…
AkashML…
↑ Sent, per 1M
9,99 DA
↓ Back, per 1M
62,24 DA
Cache read
9,99 DA
Latency
…
Throughput
…
DeepInfraUnited States…
↑ Sent, per 1M
12,31 DA
↓ Back, per 1M
56,58 DA
Cache read
none
Latency
…
Throughput
…
Crusoe…
↑ Sent, per 1M
16,64 DA
↓ Back, per 1M
83,21 DA
Cache read
16,64 DA
Latency
…
Throughput
…
Mancer 2…
↑ Sent, per 1M
18,31 DA
↓ Back, per 1M
83,21 DA
Cache read
18,31 DA
Latency
…
Throughput
…
Sail Research…
↑ Sent, per 1M
19,97 DA
↓ Back, per 1M
133,13 DA
Cache read
9,99 DA
Latency
…
Throughput
…
DigitalOcean…
↑ Sent, per 1M
19,97 DA
↓ Back, per 1M
139,79 DA
Cache read
3,99 DA
Latency
…
Throughput
…
Google…
↑ Sent, per 1M
29,96 DA
↓ Back, per 1M
119,82 DA
Cache read
none
Latency
…
Throughput
…
BaseTenUnited States…
↑ Sent, per 1M
33,28 DA
↓ Back, per 1M
166,42 DA
Cache read
33,28 DA
Latency
…
Throughput
…
BaseTen…
↑ Sent, per 1M
33,28 DA
↓ Back, per 1M
166,42 DA
Cache read
33,28 DA
Latency
…
Throughput
…
Parasail…
↑ Sent, per 1M
33,28 DA
↓ Back, per 1M
249,63 DA
Cache read
18,31 DA
Latency
…
Throughput
…
SambaNova…
↑ Sent, per 1M
46,60 DA
↓ Back, per 1M
316,19 DA
Cache read
none
Latency
…
Throughput
…
Amazon BedrockEurope…
↑ Sent, per 1M
49,93 DA
↓ Back, per 1M
199,70 DA
Cache read
none
Latency
…
Throughput
…
Nebius…
↑ Sent, per 1M
49,93 DA
↓ Back, per 1M
199,70 DA
Cache read
33,28 DA
Latency
…
Throughput
…
Amazon Bedrock…
↑ Sent, per 1M
49,93 DA
↓ Back, per 1M
199,70 DA
Cache read
none
Latency
…
Throughput
…
SiliconFlowUnited States…
↑ Sent, per 1M
49,93 DA
↓ Back, per 1M
199,70 DA
Cache read
24,96 DA
Latency
…
Throughput
…
Phala…
↑ Sent, per 1M
49,93 DA
↓ Back, per 1M
199,70 DA
Cache read
none
Latency
…
Throughput
…
Together…
↑ Sent, per 1M
49,93 DA
↓ Back, per 1M
199,70 DA
Cache read
none
Latency
…
Throughput
…
Groq…
↑ Sent, per 1M
49,93 DA
↓ Back, per 1M
199,70 DA
Cache read
24,96 DA
Latency
…
Throughput
…
DeepInfraUnited States…
↑ Sent, per 1M
66,57 DA
↓ Back, per 1M
316,19 DA
Cache read
none
Latency
…
Throughput
…
Cerebras…
↑ Sent, per 1M
116,49 DA
↓ Back, per 1M
249,63 DA
Cache read
116,49 DA
Latency
…
Throughput
…
Provider↑ Sent↓ BackCache readLatencyThroughputUptime
PrimeIntellect6,32 DA23,30 DA5,66 DA………
Near AI9,99 DA49,93 DA4,99 DA………
VeniceUnited States9,99 DA49,93 DA9,99 DA………
CoreWeaveUnited States9,99 DA56,58 DA9,99 DA………
DekaLLMAsia9,99 DA59,91 DA9,99 DA………
AkashML9,99 DA62,24 DA9,99 DA………
DeepInfraUnited States12,31 DA56,58 DAnone………
Crusoe16,64 DA83,21 DA16,64 DA………
Mancer 218,31 DA83,21 DA18,31 DA………
Sail Research19,97 DA133,13 DA9,99 DA………
DigitalOcean19,97 DA139,79 DA3,99 DA………
Google29,96 DA119,82 DAnone………
BaseTenUnited States33,28 DA166,42 DA33,28 DA………
BaseTen33,28 DA166,42 DA33,28 DA………
Parasail33,28 DA249,63 DA18,31 DA………
SambaNova46,60 DA316,19 DAnone………
Amazon BedrockEurope49,93 DA199,70 DAnone………
Nebius49,93 DA199,70 DA33,28 DA………
Amazon Bedrock49,93 DA199,70 DAnone………
SiliconFlowUnited States49,93 DA199,70 DA24,96 DA………
Phala49,93 DA199,70 DAnone………
Together49,93 DA199,70 DAnone………
Groq49,93 DA199,70 DA24,96 DA………
DeepInfraUnited States66,57 DA316,19 DAnone………
Cerebras116,49 DA249,63 DA116,49 DA………

You pay the price of the provider that served your request. Prices read 15:19 · speed over the last 30 minutes · uptime over 3 days.

See it in action

Real requests, real replies. Try any of them yourself — one click, pre-filled.

Structured output
You send

Turn this customer email into JSON with intent, urgency and order_id.

customer-email.txt
It replied

{ "intent": "refund", "urgency": "high", "order_id": "OD-58421" }

Try in Chat Lab
Writing
You send

Two lines to welcome a shop in Oran.

It replied

Welcome to our new shop in Oran — fresh bread every morning, and coffee the way the city likes it. Come in, the first cup is on us.

Try in Chat Lab

Try your own request — same model, pre-selected.

Open Chat Lab

Call it in code

Swap the base URL and the key. Your OpenAI code runs unchanged.

POST /v1/chat/completions
Base URL
https://opendunes.com/v1
Model ID
openai/gpt-oss-120b
import os
from openai import OpenAI

client = OpenAI(
    base_url="https://opendunes.com/v1",
    api_key=os.environ["OPENDUNES_API_KEY"],
)

resp = client.chat.completions.create(
    model="openai/gpt-oss-120b",
    messages=[{"role": "user", "content": "Two lines to welcome a shop in Oran."}],
)

print(resp.choices[0].message.content)
pip install openaiFull reference for this route
Sign in to run it

Runs this request once on your balance.

Parameters you can send

Standard OpenAI fields, grouped by what they control. Anything not listed is accepted and ignored.

Shaping the reply

How long, how varied, where to stop.

max_tokenstemperaturetop_ptop_kmin_ptop_astopseedfrequency_penaltypresence_penaltyrepetition_penaltylogit_biaslogprobstop_logprobs

Reasoning

Let it think before answering — and see the thinking.

reasoningreasoning_effortinclude_reasoning

Structured output

Get JSON you can parse instead of prose.

response_formatstructured_outputs

Tools

Let the model call your functions.

toolstool_choice

Questions about gpt-oss-120b

  • How much does gpt-oss-120b cost in dinars?

    About 0,017 DA per message. ↑ 5,99 DA · ↓ 22,63 DA per 1M tokens.

  • Is gpt-oss-120b free? Can I try it?

    No. You pay per use from your balance, with no subscription. Try it in Chat at the same price.

  • What is gpt-oss-120b best for?

    gpt-oss-120b reads text and writes text, keeping about 130 pages in mind at once. OpenAI released it so you can download it and run it yourself, and it is small enough for a single high-end graphics card. You choose how long it thinks (low, medium or high), you can read its full working, and it can use the tools and apps you connect, browse the web and fill in the exact form you give it.

  • gpt-oss-120b or gpt-oss-safeguard-20b?

    gpt-oss-safeguard-20b costs 0,058 DA more for the same message.

  • How do I pay for gpt-oss-120b without an international card?

    Deposit dinars with your Edahabia or CIB card. Every model, gpt-oss-120b included, is paid from that balance. No foreign card, no conversion.

  • Does gpt-oss-120b work in Arabic, French or Darja?

    OpenAI names no languages for gpt-oss-120b. Arabic, French and Darja: not stated.

  • Can it read PDFs, images or a whole book?

    It reads text. Up to 131K tokens in one request, about 130 pages. It writes up to 49 000 words at once.

  • Is gpt-oss-120b open source?

    Yes, its weights are public. Open weights on Hugging Face

  • Who makes gpt-oss-120b, and when was it released?

    OpenAI. Released 5 August 2025. It knows the world up to 30 June 2024.

Other OpenAI models, priced in dinars

See all OpenAI models
Text Embedding 3 SmallOpenAI · Oct 30, 2025

Text Embedding 3 Small turns text of up to about 8 pages into a fingerprint of its meaning, so search finds passages that mean the same thing even when the words differ. It replaces the older ada-002 and is the sensible default for search over your documents, grouping similar messages or labelling them.

8 pages per call
≈0,0067 DAper page
View model
GPT Image 2.5 SunburstOpenAI · Sep 9, 2026

GPT Image 2.5 Sunburst takes a written description, with optional reference pictures, and gives back an image, new or edited, as PNG, JPEG or WebP, with transparent backgrounds on PNG and WebP. It is the precision tier of the 2.5 series, for detailed creative work where an edit has to land exactly as asked, and it adds top quality settings named xhigh and max.

Up to 16 referencesTransparent background
≈4,38 DAper image
View model
GPT-6 Luna DecisionsOpenAI · Oct 6, 2026

GPT-6 Luna Decisions does not write text. You send content as text, a table of data or a picture, plus up to 200 named questions of three kinds: yes or no, pick among options, or grade against an ordered scale. It returns a probability for every answer, with the likeliest option or expected score, so your own system can set its thresholds for routing, moderation or marking.

1 050 pages per call
≈0,033 DAper case
View model
Text Embedding 3 LargeOpenAI · Oct 30, 2025

Text Embedding 3 Large turns text of up to about 8 pages into a fingerprint of its meaning, so search finds passages that mean the same thing even when the words differ. It is the most accurate of OpenAI's models of this kind, in English and other languages. Choose it when the quality of matches matters most, for search, grouping or labelling documents.

8 pages per call
≈0,043 DAper page
View model