Alternatives · 15 models compared · checked 17 Aug 2026
GPT-3.5 Turbo alternatives: 15 models in or near grid D1
The closest model to GPT-3.5 Turbo on the atlas is Gemma 2 27B, in the same square, at $0.65 per 1M input tokens against GPT-3.5 Turbo’s . 10 of the 15 nearest models are cheaper on input; the cheapest is Phi 4 at $0.07. We rank by position on the atlas, not by capability, because we hold no capability data.
$0.50
observed We measured
- Source
- Our catalogue
- Retrieved
- 17 Aug 2026
Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
Cite
Cite this page
ModelsAtlas. “GPT-3.5 Turbo alternatives: 15 models in or near grid D1”. Retrieved 17 Aug 2026. https://modelsatlas.com/models/openai/gpt-3-5-turbo/alternatives
Where the alternatives sit
Each square is a price band (A–H) × a context band (1–6). How the grid works
Same squareGRIDD11 modelGemma 2 27B
GPT-3.5 TurboIts 15 nearest modelsOther modelsOpen weights
View the map as a table (5 squares)
| Square | Price band | Context band | Models |
|---|---|---|---|
| GRIDB1 | <$0.10 | <32K | Phi 4 |
| GRIDC2 | $0.10–0.30 | 32–128K | Qwen2.5 7B Instruct, Voxtral Small 24B 2507, Qwen3 14B, Qwen3 30B A3B, Command R (08-2024) |
| GRIDD1 | $0.30–1 | <32K | GPT-3.5 Turbo, Gemma 2 27B |
| GRIDD2 | $0.30–1 | 32–128K | MiniMax M2-her, Nano Banana (Gemini 2.5 Flash Image), Mistral Small 3.1 24B, Qwen2.5 72B Instruct, WizardLM-2 8x22B, Qwen2.5 Coder 32B Instruct, Qwen2.5 VL 72B Instruct |
| GRIDE1 | $1–3 | <32K | GPT-3.5 Turbo Instruct |
Top 15 alternatives
Ranked by steps away on the atlas, then cheaper before pricier, then input price
| # | Model | Grid | Why it fits | Input / 1M | vs GPT-3.5 Turbo | Output / 1M | Context | |
|---|---|---|---|---|---|---|---|---|
| — | GPT-3.5 TurboOpenAI · the model you’re replacing | GRIDD1 | Baseline | $0.50 | — | $1.50 | 16K | |
| 1 | Gemma 2 27BGoogle | GRIDD1 | Same square, 30% pricier input | $0.65 | +30% | $0.65 | 8K | Compare → |
| 2 | MiniMax M2-herMiniMax | GRIDD2 | Neighbouring square D2, 40% cheaper input, 4× the context | $0.30 | −40% | $1.20 | 66K | |
| 3 | Nano Banana (Gemini 2.5 Flash Image)Google | GRIDD2 | Neighbouring square D2, 40% cheaper input, 2× the context | $0.30 | −40% | $2.50 | 33K | |
| 4 | Mistral Small 3.1 24BMistral AI | GRIDD2 | Neighbouring square D2, 30% cheaper input, 7.8× the context | $0.35 | −30% | $0.56 | 128K | Compare → |
| 5 | Qwen2.5 72B InstructQwen (Alibaba) | GRIDD2 | Neighbouring square D2, 28% cheaper input, 2× the context | $0.36 | −28% | $0.40 | 33K | Compare → |
| 6 | WizardLM-2 8x22BMicrosoft | GRIDD2 | Neighbouring square D2, 24% pricier input, 4× the context | $0.62 | +24% | $0.62 | 66K | Compare → |
| 7 | Qwen2.5 Coder 32B InstructQwen (Alibaba) | GRIDD2 | Neighbouring square D2, 32% pricier input, 2× the context | $0.66 | +32% | $1.00 | 33K | Compare → |
| 8 | Qwen2.5 VL 72B InstructQwen (Alibaba) | GRIDD2 | Neighbouring square D2, 60% pricier input, 7.8× the context | $0.80 | +60% | $1.00 | 128K | Compare → |
| 9 | GPT-3.5 Turbo InstructOpenAI | GRIDE1 | Same maker, neighbouring square E1, 200% pricier input | $1.50 | +200% | $2.00 | 4K | |
| 10 | Phi 4Microsoft | GRIDB1 | 2 squares away (B1), 86% cheaper input | $0.07 | −86% | $0.14 | 16K | |
| 11 | Qwen2.5 7B InstructQwen (Alibaba) · open weights | GRIDC2 | 2 squares away (C2), 80% cheaper input, 2× the context, open weights | $0.10 | −80% | $0.20 | 33K | |
| 12 | Voxtral Small 24B 2507Mistral AI | GRIDC2 | 2 squares away (C2), 80% cheaper input, 2× the context | $0.10 | −80% | $0.30 | 33K | |
| 13 | Qwen3 14BQwen (Alibaba) · open weights | GRIDC2 | 2 squares away (C2), 76% cheaper input, 2.5× the context, open weights | $0.12 | −76% | $0.24 | 41K | |
| 14 | Qwen3 30B A3BQwen (Alibaba) | GRIDC2 | 2 squares away (C2), 76% cheaper input, 2.5× the context | $0.12 | −76% | $0.50 | 41K | |
| 15 | Command R (08-2024)Cohere | GRIDC2 | 2 squares away (C2), 70% cheaper input, 7.8× the context | $0.15 | −70% | $0.60 | 128K |
What switching saves
3,000 input + 800 output tokens × 1,000 requests a day
Σ Derived from standard-tier prices; excludes caching and batch discounts. Use your own numbers →
Larger-context alternatives
| Model | Input / 1M | Context | |
|---|---|---|---|
| MiniMax M2-herMiniMax | $0.30 | 66K | |
| Nano Banana (Gemini 2.5 Flash Image)Google | $0.30 | 33K | |
| Mistral Small 3.1 24BMistral AI | $0.35 | 128K | Compare |
| Qwen2.5 72B InstructQwen (Alibaba) | $0.36 | 33K | Compare |
| WizardLM-2 8x22BMicrosoft | $0.62 | 66K | Compare |
Free alternatives
| Model | Input / 1M | Context | |
|---|---|---|---|
| Lyria 3 Clip PreviewGoogle | Free | 1.05M | |
| Lyria 3 Pro PreviewGoogle | Free | 1.05M |
Compare GPT-3.5 Turbo head to head
Questions
What is the best alternative to GPT-3.5 Turbo?
Gemma 2 27B ranks first: same square, 30% pricier input. Input is $0.65 per 1M tokens against $0.50 for GPT-3.5 Turbo. Ranking is by position on the atlas, not by quality.
What is the cheapest alternative to GPT-3.5 Turbo?
Phi 4 at $0.07 per 1M input tokens, 86% less than GPT-3.5 Turbo, among the 15 closest.
Is there an open-weight alternative to GPT-3.5 Turbo?
Yes, 2 of the 15: Qwen2.5 7B Instruct and Qwen3 14B.
Which alternative has a larger context window?
Mistral Small 3.1 24B, Qwen2.5 VL 72B Instruct, Command R (08-2024), MiniMax M2-her and 8 more hold more than GPT-3.5 Turbo's 16K; the largest is Mistral Small 3.1 24B at 128K.
How are alternatives ranked?
By distance on the atlas grid from GPT-3.5 Turbo's square D1: the same square first, then neighbouring squares. Within the same distance, cheaper models come before pricier ones, then the lowest input price wins. We hold no capability data, so we do not claim a model can do what GPT-3.5 Turbo does, only that it costs and holds about the same. No vendor pays for position.
More alternatives
Keep going in your AI
Opens with this page as the source. We add nothing else to the prompt.
- What is the best alternative to GPT-3.5 Turbo?
- What is the cheapest alternative to GPT-3.5 Turbo?
- Is there an open-weight alternative to GPT-3.5 Turbo?