Best cheap AI model
As of October 19, 2026, the best cheap AI model TensorFeed tracks is GLM-5.3 Flash from Z.ai (Zhipu AI), the highest-scoring budget-tier model on the TensorFeed Intelligence Index's 2026 benchmark set (62.8), at $0.15 input and $0.50 output per 1M tokens. The lowest-priced budget model is Nemotron 3.5 Lightning from NVIDIA, at $0.08 input and $0.20 output per 1M tokens. TensorFeed lists 18 budget-tier models in all.
Ordered by TensorFeed Intelligence Index score as of October 19, 2026; prices as of October 19, 2026.
Budget-tier models
TFII on the 2026 set first, then the 2024 set (ranked separately, never compared), then models the index does not rank yet.
| Model | Provider | Price in / out | TFII | Terminal-Bench 4.0 | GPQA Diamond |
|---|---|---|---|---|---|
| GLM-5.3 Flash | Z.ai (Zhipu AI) | $0.15 / $0.50 | 62.8 | 37.37% | 86.36% |
| GPT-5.6 Luna | OpenAI | $0.20 / $1.20 | 53.2 | 11.62% | 92.3% |
| MiniMax M3 | MiniMax | $0.30 / $1.20 | 29.8 | 1.01% | 92.68% |
| Gemini 3.5 Flash-Lite | $0.30 / $2.50 | 73.3 2024 set | - | 83.84% | |
| Claude Haiku 4.5 | Anthropic | $1.00 / $5.00 | 58.5 2024 set | - | 52.4% |
| Nemotron 3.5 Lightning | NVIDIA | $0.08 / $0.20 | 56.7 2024 set | - | 75.44% |
| Mistral Small | Mistral | $0.15 / $0.60 | 55.3 2024 set | - | 44.6% |
| Muse Spark 1.3 Contributor | Meta | $0.10 / $0.20 | - | - | - |
| Muse Spark 1.2 Contributor | Meta | $0.10 / $0.20 | - | - | - |
| MiMo-V2.6-Flash | Xiaomi | $0.14 / $0.28 | - | 24.24% | - |
| Claude Haiku 5.5 | Anthropic | $0.10 / $0.50 | - | 36.36% | - |
| GPT-6 Luna | OpenAI | $0.10 / $0.50 | - | 13.64% | - |
| Qwen3.8-Omni-Flash | Alibaba | $0.15 / $0.47 | - | - | - |
| Qwen3.8-Flash | Alibaba | $0.15 / $0.47 | - | - | 91.7% |
| GPT-4o-mini | OpenAI | $0.15 / $0.60 | - | - | - |
| MAI-Code-1.1-Flash | Microsoft | $0.20 / $1.20 | - | - | - |
| GLM-5.3-FlashX | Z.ai (Zhipu AI) | $0.37 / $1.25 | - | - | - |
| Gemini 3.1 Flash-Lite | $0.25 / $1.50 | - | - | 86.9% |
Prices are list prices per 1M input / output tokens. TFII is the headline TensorFeed Intelligence Index score; scores marked 2024 set come from older benchmarks and are not comparable with the rest.
How this is ranked
Cheap here means the budget tier of TensorFeed's pricing catalog: the small, fast, low-cost model each lab sells under its flagship. Within that tier, models are ordered by their headline TensorFeed Intelligence Index score, a 0 to 100 composite of public benchmarks, so the pick is the most capable model at the low end of the price range rather than simply the cheapest.
Index scores from the 2026 and 2024 benchmark sets are ranked separately and never compared, so a model scored only on the older set is listed after the 2026 set, not ranked against it. Models the index does not rank yet follow, by price.
Questions
What is the best cheap AI model?
As of October 19, 2026, the best cheap AI model TensorFeed tracks is GLM-5.3 Flash from Z.ai (Zhipu AI), the highest-scoring budget-tier model on the TensorFeed Intelligence Index's 2026 benchmark set (62.8), at $0.15 input and $0.50 output per 1M tokens. The lowest-priced budget model is Nemotron 3.5 Lightning from NVIDIA, at $0.08 input and $0.20 output per 1M tokens. TensorFeed lists 18 budget-tier models in all.
How does TensorFeed pick the best AI model for a low budget?
Cheap here means the budget tier of TensorFeed's pricing catalog: the small, fast, low-cost model each lab sells under its flagship. Within that tier, models are ordered by their headline TensorFeed Intelligence Index score, a 0 to 100 composite of public benchmarks, so the pick is the most capable model at the low end of the price range rather than simply the cheapest.
What is the cheapest AI model?
As of October 19, 2026, the lowest-priced budget-tier model TensorFeed tracks is Nemotron 3.5 Lightning from NVIDIA, at $0.08 input and $0.20 output per 1M tokens. Open-weights models have no list price; their cost is your own hosting or a provider's rate.