Skip to content
ModelCap

Price ranking

Cheapest LLM APIs

Ranked language models with a listed API price, cheapest first by a 3:1 input-weighted blend of the lowest observed provider offer. Each row shows the ModelCap Index position so cheap and capable can be read together.

Snapshot as of 19 November 2026

Ling 3.0 Tiny from InclusionAI leads the cheapest LLM API list as of 19 November 2026 with a blended price of $0.022 per 1M tokens ($0.015 in / $0.045 out), holding ModelCap Index position #149; 120 models qualify.

Models on this list
120
With openly downloadable weights
52
With a listed API price
120

Cheapest LLM APIs (top 100 of 120)

Every row is a current, canonical language model with a public position on the ModelCap Index; the list filters by a published fact and orders by a published figure.

Cheapest LLM APIs: current, canonical language models with a public ModelCap Index position and a fresh or bounded-stale listed price above zero on the OpenRouter catalogue, ordered by 0.75 × input + 0.25 × output price per million tokens
#ModelBlended $/1MModelCap IndexInput $/1MOutput $/1MContextWeights
1Ling 3.0 TinyInclusionAI$0.022#14947.2$0.015$0.045262KOpen weights
2Llama 3.1 8B InstructMeta$0.025#32513.8$0.02$0.04131KGated access
3gpt-oss-20bOpenAI$0.036#23832.2$0.018$0.09131KOpen weights
4Nex-N2.5-MiniNex AGI$0.044#4071.7$0.025$0.10262KOpen weights
5Gemma 3 4BGoogle$0.063#26726.8$0.05$0.10131KGated access
6Llama 3.2 1B InstructMeta$0.071#3674.8$0.027$0.20160KGated access
7Gemma 3 12BGoogle$0.075#20136.7$0.05$0.15131KGated access
8Granite 4.2 8BIBM Granite$0.082#26926.3$0.06$0.15131KOpen weights
9Phi 4Microsoft$0.088#30819.2$0.07$0.1416KOpen weights
10Gemma 3 27BGoogle$0.10#16042.7$0.08$0.16131KGated access
11Nemotron 3.5 LightningNVIDIA$0.103#21934.9$0.07$0.201MRestricted license
12Nemotron 3 Nano 30B A3BNVIDIA$0.105#24531.2$0.06$0.24262KRestricted license
13Mistral Small 3.2 24BMistral AI$0.106#25828.4$0.075$0.20256KOpen weights
14LFM3-1.7BLiquid$0.11#15544.8$0.08$0.20262KRestricted license
15Nex-N2.5-ProNex AGI$0.119#3274.5$0.075$0.25262KOpen weights
16Gemma 4 26B A4BGoogle$0.121#6065.9$0.076$0.255262KOpen weights
17Llama 3.2 3B InstructMeta$0.122#3588.5$0.051$0.335131KGated access
18Qwen3 32BQwen$0.13#17539.5$0.08$0.28131KOpen weights
19gpt-oss-safeguard-20bOpenAI$0.131#26028.2$0.075$0.30131KOpen weights
20Qwen3 30B A3B Instruct 2507Qwen$0.15#14747.7$0.10$0.30262KOpen weights
21Llama 4 ScoutMeta$0.15#23931.8$0.10$0.301.3MGated access
22Nemotron 3 SuperNVIDIA$0.156#16341.7$0.075$0.40262KRestricted license
23Qwen3.5-9BQwen$0.172#16441.1$0.154$0.226262KOpen weights
24MiMo-V2.6-FlashXiaomi$0.175#4668.8$0.14$0.281MOpen weights
25Nex-N2.6Nex AGI$0.19#3572.9$0.12$0.40262KOpen weights
26GPT-6 Luna ProOpenAI$0.20#5167.6$0.10$0.501MNo linked weights
27GPT-6 LunaOpenAI$0.20#5866.3$0.10$0.501MNo linked weights
28Muse EmberMeta$0.20#6365.2$0.10$0.501MNo linked weights
29Claude Haiku 5.5Anthropic$0.20#9358.3$0.10$0.501MNo linked weights
30LFM2.5-24B-A2BLiquid$0.212#13949.4$0.05$0.70128KRestricted license
31Qwen3 30B A3BQwen$0.215#22234.6$0.12$0.5041KOpen weights
32Qwen3.8 FlashQwen$0.23#10556.1$0.15$0.471MRestricted license
33GLM 5.3 FlashZ.ai$0.237#2676.4$0.15$0.501MOpen weights
34Nemotron 3.5 Super VLNVIDIA$0.237#13350.9$0.15$0.501MRestricted license
35Gemma 4 31BGoogle$0.25#4270.7$0.20$0.40262KOpen weights
36MiniMax M3.1 FlashMiniMax$0.262#3374.1$0.15$0.601MRestricted license
37DeepSeek V4.1 Flash 1017DeepSeek$0.262#9258.5$0.15$0.601MOpen weights
38Qwen3 Coder 30B A3B InstructQwen$0.262#13051.2$0.15$0.60262KOpen weights
39Solar Pro 4Upstage$0.262#14548.0$0.15$0.60524KNo linked weights
40gpt-oss-120bOpenAI$0.262#17439.8$0.15$0.60131KOpen weights
41Qwen3 VL 30B A3B InstructQwen$0.262#18438.7$0.15$0.60262KOpen weights
42GPT-4o-miniOpenAI$0.262#22434.4$0.15$0.60128KNo linked weights
43Mistral Small 4Mistral AI$0.262#27924.6$0.15$0.60262KOpen weights
44Command R (08-2024)Cohere$0.262#31018.4$0.15$0.60128KGated access
45Qwen3 Coder NextQwen$0.29#8659.5$0.12$0.80262KOpen weights
46DeepSeek V3.2DeepSeek$0.29#8958.8$0.26$0.38164KOpen weights
47Qwen3.6 35B A3BQwen$0.30#11652.6$0.10$0.90262KOpen weights
48Qwen3 235B A22B Instruct 2507Qwen$0.305#9158.5$0.14$0.80262KOpen weights
49GLM 4.5 AirZ.ai$0.31#15445.1$0.13$0.85131KOpen weights
50Qwen3 Next 80B A3B InstructQwen$0.35#12451.7$0.10$1.10262KOpen weights
51Qwen2.5 72B InstructQwen$0.37#25928.3$0.36$0.4033KRestricted license
52Qwen3 VL 8B InstructQwen$0.375#22134.7$0.25$0.75262KOpen weights
53GLM 5.3Z.ai$0.402#1978.5$0.07$1.401MRestricted license
54Beam 1.0 501B A23BReflection$0.438#6763.4$0.25$1.001MOpen weights
55Step 3.7 FlashStepFun$0.438#12851.6$0.20$1.15262KOpen weights
56GLM 4.6VZ.ai$0.45#15146.1$0.30$0.90131KOpen weights
57DeepSeek V3.1 TerminusDeepSeek$0.453#9756.9$0.27$1.00164KOpen weights
58GPT-5.4 NanoOpenAI$0.463#12651.6$0.20$1.25400KNo linked weights
59Inkling 2 SmallThinking Machines$0.475#9856.8$0.30$1.001MOpen weights
60Qwen3 Coder 480B A35BQwen$0.475#14248.9$0.30$1.00262KOpen weights
61Llama 4 MaverickMeta$0.512#22633.5$0.35$1.00524KGated access
62LongCat 2.0Meituan$0.525#2377.6$0.30$1.201MOpen weights
63MiniMax M3MiniMax$0.525#6663.7$0.30$1.201MRestricted license
64Motif 3Motif Technologies$0.525#8259.9$0.30$1.20262KOpen weights
65MiMo-V2.6-ProXiaomi$0.544#2078.4$0.435$0.871MOpen weights
66Qwen3.7 PlusQwen$0.56#4370.6$0.32$1.281MNo linked weights
67Qwen3.8 27BQwen$0.595#6863.1$0.06$2.201MOpen weights
68Command A+Cohere$0.60#22334.6$0.30$1.50192KOpen weights
69Qwen3 VL 235B A22B InstructQwen$0.632#10755.6$0.21$1.90262KOpen weights
70Muse Glimmer 30BMeta$0.637#5666.8$0.35$1.50131KOpen weights
71Llama 3.3 70B InstructMeta$0.71#24431.4$0.71$0.71131KGated access
72Qwen3.5-122B-A10BQwen$0.715#9956.7$0.26$2.08262KOpen weights
73Qwen2.5 Coder 32B InstructQwen$0.745#29023.3$0.66$1.0033KOpen weights
74MiMo-V2.5Xiaomi$0.80#6564.4$0.40$2.001MOpen weights
75Gemini 3.5 Flash LiteGoogle$0.85#4171.1$0.30$2.501MNo linked weights
76Nemotron 3 UltraNVIDIA$0.925#26128.0$0.50$2.20262KRestricted license
77R1 0528DeepSeek$1.00#9558.1$0.571$2.29164KOpen weights
78Seed 2.1 TurboByteDance Seed$1.00#2876.0$0.50$2.50262KNo linked weights
79Mistral Large 4Mistral AI$1.05#9658.1$0.70$2.101MRestricted license
80Kimi K2 ThinkingMoonshot AI$1.07#7162.3$0.60$2.50262KRestricted license
81Qwen3.5 397B A17BQwen$1.09#5267.2$0.45$3.00262KOpen weights
82DeepSeek V4 Pro 0813DeepSeek$1.19#3473.0$0.111$4.441MOpen weights
83Qwen3 235B A22B Thinking 2507Qwen$1.21#11852.3$0.45$3.50128KOpen weights
84Hy4 previewTencent$1.25#4569.5$0.834$2.501MOpen weights
85Step 5 PreviewStepFun$1.43#4968.4$1.00$2.701MRestricted license
86Grok 4.20 Multi-AgentxAI$1.56#2776.2$1.25$2.502MNo linked weights
87Gemini 3.8 FlashGoogle$1.65#881.4$0.825$4.131MNo linked weights
88GPT-5.4 MiniOpenAI$1.69#4768.5$0.75$4.50400KNo linked weights
89InklingThinking Machines$1.76#6265.6$1.00$4.05524KOpen weights
90Muse Spark 1.3Meta$2.00#782.4$1.25$4.251MNo linked weights
91Muse Spark OpenMeta$2.00#2178.4$1.25$4.251MGated access
92GPT-6.1 TerraOpenAI$2.00#3772.5$1.00$5.001MNo linked weights
93Qwen4 MaxQwen$2.14#3672.8$1.42$4.271MNo linked weights
94Kimi K3.1 ProMoonshot AI$2.39#1180.0$0.253$8.801MRestricted license
95GLM 5.3 PrimeZ.ai$2.58#1279.7$1.68$5.281MNo linked weights
96Gemini 4 HydrogenGoogle$3.00#980.4$1.50$7.501MNo linked weights
97Grok 4.7xAI$3.00#5966.1$2.00$6.00500KNo linked weights
98Mistral Medium 3.5Mistral AI$3.00#8759.4$1.50$7.50262KRestricted license
99Mixtral 8x22B InstructMistral AI$3.00#31815.9$2.00$6.0066KOpen weights
100o3OpenAI$3.50#7561.0$2.00$8.00200KNo linked weights

Cheapest LLM APIs: common questions

Which model tops the cheapest LLM API list right now?

Ling 3.0 Tiny (InclusionAI), as of 19 November 2026, with a blended price of $0.022 per 1M tokens ($0.015 in / $0.045 out) and ModelCap Index position #149.

How is the cheapest LLM API list built?

It contains current, canonical language models with a public ModelCap Index position and a fresh or bounded-stale listed price above zero on the OpenRouter catalogue, ordered by 0.75 × input + 0.25 × output price per million tokens. Nothing is estimated for the list itself: the filter is a published fact and the order is a published figure, and every row links to the model page that shows its evidence.

How many models qualify for the cheapest LLM API list?

120 models qualify; the page shows the first 100. The population is the live ModelCap Index, so a model appears here the moment it holds a public position and meets the filter.

Which model on this list ranks highest on the ModelCap Index?

Gemini 4 Carbon holds ModelCap Index position #1, the highest of any model in the list, and sits at #101 on the full list, beyond the rows shown here by blended $/1m.

What is the cheapest model among the top ten on the ModelCap Index?

Gemini 3.8 Flash (Index #8) at $1.65 blended per 1M tokens — $0.825 input, $4.13 output.

Why order by a blended price instead of the output price?

Most production traffic is prompt-heavy, so the list weights input at 75% and output at 25%; both raw prices are shown in the table so a generation-heavy workload can re-order by the output column on the model pages.

Which is the best open-weight model on this list?

Ling 3.0 Tiny from InclusionAI is the highest-placed model with openly downloadable weights (#1 here, ModelCap Index #149).

How current is the cheapest LLM API list?

Positions, prices and context figures come from the same published snapshot dated 19 November 2026. Publication and page updates happen separately; use the snapshot date shown on the page.

Explore further