GPT-3.5 Turbo (older v0613)
openai/gpt-3.5-turbo-0613- OpenAI
- Family · GPT-3 Turbo
- Released 13 Jun 2023
- Text → Text
- Tools · JSON
Not on OSMEXO yet
Spec & list price
Context4K
Max output4K
Input / 1M$1.00
Output / 1M$2.002× input
Usage
Tokens / day17.8M▲ 680% 7d
Usage rank#379
Requests / day31.3K
Tool calls / req0.01
Per request · 30 days
Avg input2920.0× median
Avg output810.1× median
Cache hit4.3%0.1× median
Tool-call failures0.2%0.1× median
Consumption
Peak 128.6M 28 OctLow 827.9K 17 OctRank today #379
News & trends
Trending · 14 days · GPT-3 Turbo family
- OpenAI apaga versiones antiguas de GPT-4 y GPT-3.5 Turbo en su API: si tu app las usa, toca migrar30 views
- gpt-3.5-turbo deja de funcionar el 23 de octubre: qué modelo usar15 views
- GPT-3.5 Turbo will be no more in three days26 upvotes
News · latest from OpenAI
- GPT-6.1 Terra vs. Claude Opus 5.5: Cheaper and faster in three tests at maxthenewstack.io · 2 Nov
- Chinese labs now target OpenAI's GPT-6 Astra for distillationfork.zone · 1 Nov
- GPT-6.1 Terrasimonwillison.net · 31 Oct
- GPT-5.6 Sol can beat Factorio on peaceful moderyanmadden.net · 31 Oct
Speed
Output speed#75/33494 tok/s1.5× faster than median
First token#760.6 s1.8× quicker than median
Technical details
- Tokenizer
- GPT
- Knowledge cutoff
- 2021-09-30
- Moderation
- none
Request parameters 14
- tools
- tool_choice
- response_format
- structured_outputs
- temperature
- top_p
- max_completion_tokens
- stop
- seed
- frequency_penalty
+4 more
- logit_bias
- logprobs
- presence_penalty
- top_logprobs
Interesting facts
- Released 1238 days ago, the newest of 2 GPT-3 Turbo models.
- Weekdays average 15.0M tokens a day against 7.7M on weekends.
When to use
Best for
- Maintaining legacy integrations pinned to this version
- Short tool-calling prompts
- Simple low-cost text tasks
Think twice for
- Long documents because context is tiny
- New projects
- Image input
- Long outputs
Same family
GPT-3 Turbo- 5.5
Want GPT-3.5 Turbo (older v0613) on OSMEXO?
Sign in to request it. Requests are reviewed; a request does not guarantee availability.