The June 2023 snapshot introduced function calling to GPT-3.5 Turbo, enabling structured JSON output for tool use through a new functions API parameter. It also improved system message steerability over the original 0301 snapshot, making it the first 3.5 variant widely adopted for production API workflows requiring consistent instruction-following.
Discover the power and diversity of large language models available with Telnyx. Explore the options below to find the perfect model for your project.
| Organization | Model Name | Tasks | Languages Supported | Context Length | Parameters | Model Tier | License |
|---|---|---|---|---|---|---|---|
| No data available at this time, please try again later. |
Powered by our own GPU infrastructure, select a large language model, add a prompt, and chat away. For unlimited chats, sign up for a free account on our Mission Control Portal here.
GPT-3.5 Turbo 0613 was deprecated by OpenAI in favor of newer snapshots. The 0125 variant is the recommended replacement, and OpenAI suggests migrating to GPT-4o mini for new projects.
GPT-3.5 Turbo 0613 was the first snapshot to support function calling, making it a foundational model for tool-using applications. While surpassed by newer models, it remains functional for basic chat and classification tasks at low cost.
GPT-3.5 Turbo added chat optimization, function calling, and significantly lower pricing compared to GPT-3's completion-based API. The was the first to introduce function calling, enabling structured tool integration.
GPT-3.5 Turbo 0613 shares the 70.0% MMLU (5-shot) and 7.94 MT-Bench baseline of the GPT-3.5 Turbo family. As the first snapshot to introduce function calling, its quality differentiation from later snapshots (1106, 0125) is in structured output reliability rather than raw benchmark performance. Compared to Mixtral 8x7B (70.6% MMLU, 8.30 MT-Bench) on the sheet, it trails by a narrow margin on both measures.
The cost per 1,000 tokens for running the model with Telnyx Inference is $0.0010. To illustrate, if a marketing ops team were to analyze 1,000,000 customer chats, assuming each chat is 1,000 tokens long, the total cost would be $1,000.
GPT-4 0613 is a separate model, the June 2023 snapshot of GPT-4 with function calling support. It is not related to GPT-3.5 Turbo 0613 beyond sharing the same release date convention. Both are available through OpenAI's API.
GPT-3.5 Turbo 0613 is priced at $1.50 per million input tokens and $2.00 per million output tokens. The newer 0125 snapshot offers the same capability at lower pricing.
