Best low-cost AI models

190 models · Lowest output price. Ranked models only. Method

Ranked models

November 2026

#ModelOutput / 1MIndexSpeed

01dots3-note PreviewFree63.4—

01Apodex 1.1 MiniFree55.7—

01Nemotron 3 UltraFree50.8—

01Gemma 4 31BFree47.935/s

01Gemma 4 26B A4BFree44.5—

01North Mini CodeFree38.070/s

07Ling 3.0 Flash$0.06350.973/s

08GLM-5.3-Flash$0.07863.1122/s

09GPT-OSS 20B$0.1341.7310/s

10Ling 3.1 Flash$0.13567.281/s

11Phi-4$0.1422.493/s

12Mercury 2.5$0.1546.2125/s

12Granite 4.2 8B$0.1537.6101/s

14LFM3-1.7B$0.2041.5218/s

15MiMo-V2.6-Flash$0.2864.328/s

16DeepSeek V4 Flash 0731$0.3061.961/s

16Llama 4 Scout$0.3026.863/s

18Mercury 3$0.3552.3295/s

19DeepSeek V3.2$0.3844.219/s

20DeepSeek V4.1 Flash 1017$0.4068.5261/s

20GPT-5 nano$0.4041.981/s

20Gemini 2.5 Flash-Lite$0.4027.1140/s

20GPT-4.1 nano$0.4026.356/s

24Qwen3.8-Omni-Flash$0.4762.741/s

25Claude Haiku 5.5$0.5066.3159/s

25GPT-6 Luna$0.5065.566/s

25Muse Ember$0.5064.8158/s

25Ternary Bonsai 2 27B$0.5052.221/s

25Grok 3 Mini$0.5044.365/s

25Grok 4.1 Fast$0.5033.3138/s

25LFM2.5-2.6B$0.5029.3189/s

32MiniMax M3.1 Flash$0.6069.3100/s

32Ripple-65B Preview$0.6052.363/s

32Hy3 Preview$0.6052.369/s

32Solar Pro 4$0.6052.144/s

32GPT-OSS 120B$0.6043.434/s

32Solar Pro 3$0.6039.2114/s

32Mistral Small 4$0.6037.9114/s

32GPT-4o mini$0.6026.256/s

40LFM2.5-24B-A2B$0.7039.8280/s

40Celeris-1$0.7033.71,555/s

42Trinity-Large-Thinking$0.8043.5138/s

42Qwen3 235B 2507$0.8042.348/s

44MiniMax M2.7$0.8451.235/s

45MiMo-V2.6-Pro$0.8770.231/s

46Qwen3.6-35B-A3B$0.9050.665/s

47Inkling 2 Small$1.0066.4—

47Qwen3.5-35B-A3B$1.0049.443/s

47Trinity-Large-Preview$1.0042.2—

47Llama 4 Maverick$1.0034.461/s

51MiniMax M2.5$1.0850.244/s

52DeepSeek V3$1.1029.937/s

53Step 3.7 Flash$1.1555.3112/s

54GPT-5.6 Luna$1.2065.551/s

54MiniMax M3$1.2055.779/s

54Inkling-Small$1.2055.180/s

54Muse Glimmer 30B$1.2053.090/s

54MiniMax M2.1$1.2049.134/s

59Hy4$1.2572.052/s

59GPT-5.4 nano$1.2550.991/s

59Claude 3 Haiku$1.2522.073/s

62Qwen3.7 Plus$1.2857.751/s

63DeepSeek V4 Flash Vision Exp$1.3260.474/s

64GLM-5.3$1.4065.886/s

65Gemini 3.1 Flash-Lite$1.5051.9134/s

65Gemini 3.1 Flash Lite Preview$1.5049.6151/s

65Grok Code Fast 1$1.5041.2172/s

65Command A+$1.5040.2161/s

65Mistral Large 3$1.5034.361/s

70GPT-4.1 mini$1.6034.572/s

71DeepSeek V4 Pro 0423$1.7857.689/s

72MiMo-V2.5-Pro$1.8056.347/s

73Qwen3 235B-A22B$1.8238.144/s

74Qwen3.8-27B$1.8859.2113/s

75Qwen3.6 Plus$1.9554.644/s

76MiMo-V2.5$2.0054.84/s

76Grok Build 0.1$2.0052.7147/s

76GPT-5 mini$2.0048.478/s

76GLM-4.6$2.0041.830/s

76Mistral Medium 3$2.0027.194/s

81Qwen3.5-122B-A10B$2.0851.970/s

82Mistral Large 4$2.1060.7113/s

83DeepSeek-R1-0528$2.1544.620/s

84DeepSeek V3.2 (Thinking)$2.1949.4—

85GLM-4.7$2.2050.028/s

85GLM-4.5$2.2045.725/s

87Kimi K2$2.3039.637/s

88Hy4 preview$2.5066.467/s

88Seed 2.1 Turbo$2.5057.437/s

88Grok 4.3$2.5053.361/s

88Grok 4.20$2.5052.872/s

88Gemini 3.5 Flash-Lite$2.5052.0130/s

88Kimi K2 Thinking$2.5048.248/s

88DeepSeek-R1$2.5038.918/s

88Gemini 2.5 Flash$2.5035.095/s

96Qwen3.5-27B$2.6051.369/s

97Step 5 Preview$2.7068.4105/s

98Gemini 3 Flash Preview$3.0057.578/s

98Gemini 3 Flash$3.0055.0198/s

98Kimi K2.5 (Reasoning)$3.0053.0—

101Qwen3.6-27B$3.2052.160/s

101GLM-5$3.2051.645/s

103Kimi K2.5$3.3252.236/s

104Kimi K2.6$3.4158.281/s

105Kimi K2.7 Code$3.5057.735/s

105Qwen3 235B 2507 (Reasoning)$3.5047.4108/s

107Qwen3.5 397B (Reasoning)$3.6055.3—

107Qwen3.5 397B$3.6053.859/s

109Gemini 3.7 Flash$3.7568.585/s

109Gemini 3.6 Flash$3.7563.7121/s

111Qwen3 Max$3.9045.062/s

112GLM-5-Turbo$4.0051.925/s

112GLM-5V-Turbo$4.0048.130/s

114Inkling$4.0554.885/s

115Gemini 3.8 Flash$4.1367.868/s

116Muse Spark 1.3$4.2570.176/s

116Muse Spark Open$4.2565.186/s

116Muse Spark 1.2$4.2564.9123/s

116Muse Spark 1.1$4.2564.6152/s

120GLM-5.1$4.4056.0102/s

120o4-mini$4.4044.7122/s

120o3-mini$4.4040.0173/s

120o1-mini$4.4033.5—

124Qwen3.7 Max$4.4360.568/s

125DeepSeek V4 Pro 0813$4.4465.542/s

126GPT-5.4 mini$4.5054.595/s

127GPT-6.1 Terra$5.0077.146/s

127Claude Haiku 4.5$5.0043.288/s

129Grok 4.7$6.0069.151/s

129Grok 4.6$6.0067.856/s

129Grok 4.5$6.0063.753/s

129Qwen3.8 Max (0902)$6.0063.433/s

129Mistral Large 2$6.0027.470/s

134Qwen 3.6 Max (preview)$6.1657.275/s

135GLM-5.2$6.8060.6111/s

136Mistral Medium 3.5 128B$7.5050.3—

136Mistral Medium 3.5$7.5042.792/s

138o3$8.0051.865/s

138GPT-4.1$8.0038.074/s

140Gemini 3.5 Flash$9.0064.2137/s

141Gemini 4 Carbon$10.0082.731/s

141Gemini 4 Argon$10.0078.528/s

141GPT-6.1 Sol$10.0078.034/s

141Claude Sonnet 5.5$10.0077.2107/s

141Kimi K3.1 Pro$10.0075.743/s

141GPT-6 Sol$10.0072.861/s

141Claude Sonnet 5$10.0064.275/s

141GPT-5.1$10.0055.350/s

141GPT-5$10.0055.076/s

141GPT-5.1-Codex-Max$10.0053.855/s

141GPT-5 (high)$10.0053.676/s

141GPT-5 (medium)$10.0053.076/s

141GPT-5.1-Codex$10.0052.955/s

141Gemini 2.5 Pro$10.0046.691/s

141GPT-4o$10.0028.766/s

156GPT-5.6 Terra$12.0068.966/s

156Gemini 3.1 Pro$12.0062.3151/s

156Gemini 3.1 Pro Preview$12.0061.577/s

156Gemini 3 Pro$12.0057.9109/s

160GPT-5.3 Codex$14.0062.052/s

160GPT-5.2-Codex$14.0059.944/s

160GPT-5.2$14.0059.045/s

163GPT-5.4$15.0064.838/s

163Claude Sonnet 4.6$15.0057.044/s

163Claude Sonnet 4.5$15.0047.945/s

163Claude Sonnet 4$15.0043.049/s

163Claude 4 Sonnet$15.0039.540/s

163Claude 3.5 Sonnet$15.0032.2—

169Kimi K3$16.0066.944/s

170Claude Opus 5.5$20.0078.676/s

170Muse Wonder$20.0073.180/s

172GPT-5.6 Sol$22.0073.064/s

173Claude Opus 5$25.0073.848/s

173Claude Opus 4.8$25.0067.769/s

173Claude Opus 4.7 (Adaptive)$25.0063.2—

173Claude Opus 4.7$25.0062.047/s

173Claude Opus 4.6$25.0058.036/s

173Claude Opus 4.5$25.0052.039/s

179GPT-5.5$30.0069.152/s

180Claude Fable 5.2$40.0085.664/s

181GPT-6 Astra$50.0078.034/s

181Claude Fable 5.1$50.0076.269/s

181Claude Fable 5$50.0074.056/s

181Claude Mythos 5$50.0070.8—

185o1$60.0044.597/s

186Claude Opus 4.1$75.0047.512/s

186Claude 3 Opus$75.0023.3—

188GPT-4.5$150.0038.8—

189GPT-5.5 Pro$180.0074.819/s

189GPT-5.4 Pro$180.0073.321/s