SmophyAI

Published daily · Next update 02:00 UTC

Price-band ranking

Best AI models Under $0.50 / 1M tokens

As of August 9, 2026, Z.ai: GLM 5.2 (batch) is the top-quality model priced under $0.50 / 1m tokens, scoring 52.6 on the Artificial Analysis Intelligence Index among 30 eligible models in this price tier.

  1. #1 Z.ai: GLM 5.2 (batch) - Quality 52.6 · $0.11/1M
  2. #2 OpenAI: GPT-5.6 Luna (batch) - Quality 52.3 · $0.22/1M
  3. #3 DeepSeek: DeepSeek V4 Flash 0423 - Quality 51.8 · $0.11/1M
  4. #4 Tencent: Hy3 preview - Quality 42.2 · $0.10/1M
  5. #5 Nex AGI: Nex-N2-Pro - Quality 41.7 · $0.44/1M
  6. #6 OpenAI: GPT-5.4 Nano (batch) - Quality 39.7 · $0.46/1M
  7. #7 Xiaomi: MiMo-V2.5 - Quality 38 · $0.18/1M
  8. #8 Ling-3.0-flash - Quality 37.8 · $0.03/1M
30 models priced under $0.50 / 1m tokens
#ModelQualityPrice/1M
01Z.ai: GLM 5.2 (batch)52.6$0.11
02OpenAI: GPT-5.6 Luna (batch)52.3$0.22
03DeepSeek: DeepSeek V4 Flash 042351.8$0.11
04Tencent: Hy3 preview42.2$0.10
05Nex AGI: Nex-N2-Pro41.7$0.44
06OpenAI: GPT-5.4 Nano (batch)39.7$0.46
07Xiaomi: MiMo-V2.538$0.18
08Ling-3.0-flash37.8$0.03
09DeepSeek: DeepSeek V3.232.6$0.30
10Qwen: Qwen3.6 35B A3B32.1$0.36
11inclusionAI: Ring-2.6-1T31.1$0.21
12StepFun: Step 3.7 Flash30.9$0.44
13DeepSeek: DeepSeek V3.1 Terminus30.6$0.45
14Google: Gemma 4 31B (free)29.7$0.16
15Z.ai: GLM 4.629.3$0.45
16Google: Gemma 4 26B A4B (free)26.1$0.14
17NVIDIA: Nemotron 3 Super (free)25.7$0.45
18Qwen: Qwen3.5-35B-A3B24.3$0.35
19OpenAI: gpt-oss-120b24.1$0.07
20Inception: Mercury 221.9$0.38
21Qwen: Qwen3.5-9B21.8$0.11
22Qwen: Qwen3 Coder Next21.3$0.29
23Mistral: Mistral Small 419.7$0.26
24Arcee AI: Trinity Large Thinking18.6$0.38
25OpenAI: gpt-oss-20b (free)15.2$0.05
26DeepSeek: DeepSeek V3 032415.2$0.48
27Upstage: Solar Pro 314.5$0.26
28NVIDIA: Nemotron 3 Nano 30B A3B (free)14.5$0.09
29inclusionAI: Ling-2.6-flash14.2$0.02
30Mistral: Ministral 3 14B 251211.2$0.20

Source: OpenRouter (openrouter.ai/rankings), as of August 9, 2026.

Benchmark scores: Artificial Analysis (artificialanalysis.ai) via OpenRouter (openrouter.ai/rankings).

Methodology: smophy.ai/benchmark/methodology

Token counts originate from each provider's own tokenizer and are not directly comparable across providers.