LeadsCalc
Interactive leaderboard

Cheapest LLM APIs 2026: Low-Cost AI Models Ranked by Workload

Discover the cheapest LLM APIs in 2026: blended input/output cost, batch discounts, and prompt caching. Compare budget AI models for startups and agencies in the US, Canada, and Australia.

Lowest estimated monthly API cost for the same workload in 2026

The cheapest tab ranks models by estimated spend for your exact monthly requests and token pattern, including batch and cached pricing only when our database confirms eligibility. Founders and agencies in the United States, Canada, and Australia use it to protect margins on high-volume chat, summarization, and RAG pipelines without guessing list prices from blog posts.

Est. monthlyROI scoreCodingReasoningSpeedMathContextOverallOpen-weight

Workload & pricing toggles

Workload presets

Same three scenarios as the main AI API calculator: moderate traffic, large RAG-style context, or per-request max tokens with a lower request count.

Include Vision / Image Processing

Off — no image fees in cost estimates for vision-capable models.

Turn On to include image fees.

OffOn

Use Cached Pricing

Enable to get 50% off input tokens where cached rates apply

OffOn

Deep Reasoning / Thinking Mode

Model hidden reasoning / extended thinking charged like output tokens when enabled.

OffOn

Batch Pricing

Enable for 50% off input & output where batch/async pricing applies

OffOn
≈ $100.00/mo
8K
1K—1.0M
≈ $100.00/mo
2K
100—500K
≈ $200.00 total
5K
10—100K

Cached / batch est. monthly values only change after the pipeline sets supports_caching or supports_batch in Supabase. The toggles here narrow the table to models whose catalog or provider typically supports those modes.

Magic quadrant (top 15)

X: est. monthly · Y: Cheapest (est. monthly) · Dot: provider color · Hover for rank, model & details

Full leaderboard

Showing 48 of 513 models.

PickModelEst. monthlyROI scoreCodingReasoningSpeedMathContextOverall
Ox Alpha
Free
79
90
88
85
85
1.0M
88Primary source: Web digest (102.5 tok/s). Lacking exact SWE-bench or GPQA percentages for Ox Alpha, capabilities are inferred for a flagship reasoning tier. Speed mapped to 85 based on the cited high throughput.
TypeSafe: Jev Router
VARIABLE
54
40
45
100
40
1.0M
43No public SWE-bench or GPQA scores exist; LMMarketCap reports 40/100 for coding. Mapped logic and math to ~40 as a fast System One router. Speed is 100 due to 250k tokens/s and 70ms latency.
Union Alpha
Free
57
50
50
50
50
262K
50
Elephant
Free
77
90
83
70
88
262K
86
Owl Alpha
Free
66
65
68
85
60
1.0M
65No exact scores for Owl Alpha; inferred as a lightweight reasoning model ('fewer parameters', 'designed for speed'). Mapped to mid-tier 0-100 scale (Coding 65, Logic 65) reflecting its agentic focus but smaller size.
Space Bunny Alpha
Free
80
92
89
90
94
1.0M
91SWE-bench Verified at 65.4% maps to 92 coding. MMLU-Pro at 86.2% maps to 90 logic. MATH-500 at 92.8% maps to 94 math. High 118.5 TPS maps to 90 speed. No tier adjustment needed for this stealth flagship.
NVIDIA: Switchyard
VARIABLE
80
90
90
70
95
1.0M
91Switchyard is a router. Evidence cites top models hitting HumanEval 95-99%, MMLU 88-94%, and GPQA 65-74%, mapped to ~90s for coding and logic. NVIDIA NIM evidence shows 'detailed thinking' reasoning support. No vision or caching cited.
Auto Router (Beta)
VARIABLE
77
85
85
70
85
2.0M
85Primary source: OpenRouter docs (no raw scores). Inferred frontier-level capabilities (~85) across coding, logic, instruction, and math based on task-aware routing to popular models. Speed estimated at 70. Reasoning explicitly omitted for dynamic routers.
Google: Lyria 3 Pro Preview
Free
67
70
70
55
60
1.0M
68Evidence notes Lyria 3 Pro scores well on SWE-bench and MMLU without exact figures. Mapped to 70s for Pro tier. Speed is 39.5 tok/s (55). Multimodal audio generation from images supported; default Pro vision price applied.
OpenRouter: Fusion
VARIABLE
77
85
85
40
85
1.0M
85No explicit Fusion scores provided. Inferred as a heavyweight ensemble ('panel of expert models'), mapping to ~85 across coding (SWE-bench) and logic (GPQA). Speed is rated lower (40) due to multi-model deliberation and web search overhead.
Auto Router
VARIABLE
80
90
90
70
90
2.0M
90Auto Router optimizes across models for best output. Evidence cites top models reaching 92.3% MMLU. Mapped to 90 across logic, coding, and math to reflect frontier routing capabilities. Vision price defaulted to $0.007 per tier guidelines.
Body Builder (beta)
VARIABLE
54
45
43
90
40
128K
43Lacking specific benchmarks, inferred from cited 1B-scale Fast-dLLM v2 (43.5 avg across HumanEval, GPQA, GSM8K). Mapped to 40-45 for coding, logic, math. Speed rated 90 for lightweight specialized API tool.
Free Models Router
Free
55
45
48
90
45
200K
46No specific benchmarks cited for the Free Router. Inferred lightweight tier scores (Coding/Logic ~45) based on typical free 8B-class models. Speed rated high (90). Vision price is $0 as the endpoint is free.
Google: Lyria 3 Clip Preview
Free
31
0
5
50
0
1.0M
3Lyria 3 is a specialized music generation model lacking standard LLM benchmarks (SWE-bench, GPQA). Assigned 0 for coding/logic/math. Speed mapped to 50 from 38 tok/s. Multimodal scored 85 for native image-to-audio generation.
Pareto Code Router
VARIABLE
77
88
85
70
85
2.0M
86OpenRouter docs state this is a router defaulting to High tier coding models based on Artificial Analysis percentiles. Lacking specific raw benchmarks, scores are mapped to ~85 reflecting flagship-level routed performance. Text-only inputs confirmed.
inclusionAI: Ling-2.6-flash$0.7082
65
68
90
65
262K
66Evidence cites GPQA, AIME, and LiveCodeBench without raw scores. Mapped to ~65 for coding/logic based on claimed ~40B dense equivalence. Speed scored 90 due to 200+ tokens/s. Flash tier adjustment applied.
Mistral: Mistral Nemo$1.4661
45
50
90
35
131K
45GPQA 5.37% maps to Logic 35. IFEval 63.8% maps to Instruction 65. MATH Lvl 5 is 12.69%. As a 12B lightweight model, it scores lower on coding/logic than flagships but achieves high speed (90).
inclusionAI: Ling 3.0 Flash$1.4766
35
63
95
55
262K
54Evidence lacks exact Ling-3.0-flash scores, so inferred from Ling-2.6-flash (GPQA 59.3%, Coding 25.3) and Flash tier traits. Mapped GPQA to Logic 60, Coding to 35. Speed set to 95 reflecting 94 tok/s lightweight MoE architecture.
ScaleDown: Extract$1.6063
40
63
95
40
1.0M
51No specific SWE-bench or GPQA scores cited for ScaleDown Extract. Inferred coding (40) and logic (50) based on its small, task-specific extraction tier. Speed (95) reflects '20X faster' claims.
ScaleDown: Summarize$1.6053
20
45
85
20
1.0M
33No benchmarks provided for ScaleDown Summarize. Inferred as a specialized lightweight summarization model. Assigned conservative scores for coding and math, with moderate instruction and high speed due to its input-only, text-only architecture.
OpenAI: gpt-oss-20b$1.6280
70
80
90
95
131K
81GPQA 71.5% and MMLU 85.3% map to Logic 80. AIME 2025 98.7 maps to Math 95. As a 21B lightweight MoE, Speed is 90. Coding inferred at 70 due to lack of explicit SWE-bench.
IBM: Granite 4.0 Micro$1.8065
45
58
85
65
131K
56Based on GPQA 32.15% (Logic ~35) and HumanEval 81.00% (Coding ~45). IFEval averages 84.32% (Instruction ~80). As a 3B 'Micro' tier model, scores reflect lower reasoning capacity compared to flagships, but strong instruction following.
Nex AGI: Nex-N2-Mini$2.0068
45
70
90
70
262K
64No exact N2-Mini benchmarks provided; inferred from N1 8B/32B SWE-bench Verified (20.3-50.5) and N2 Pro MMLU (89.2). Mapped coding to 45 and logic to 70, adjusting downward for the Mini tier. Speed rated high (90).
Nex AGI: Nex-N2.5-Mini$2.0077
74
86
90
75
262K
80SWE-bench Verified at 74.4 maps to 74 coding. GPQA Diamond at 82.6 maps to 82 logic. IFEval at 89.1 maps to 89 instruction. As a 'Mini' tier, speed is rated high (90), with strong multimodal scores (OmniDoc 89.7).
Google: Gemma 3 4B$2.1067
45
68
95
75
131K
64Based on IFEval (90.2%) mapped to 90 Instruction, MATH (75.6%) to 75 Math, and MMLU-Pro (43.6%) to 45 Logic. As a 4B lightweight tier, Coding (MBPP 63.2%) maps to 45, while Speed is rated 95.
Sao10K: Llama 3 8B Lunaris$2.1062
45
63
90
45
8K
54Evidence lacks specific benchmark scores. Inferred from Llama 3 8B lightweight tier: coding and math mapped to ~45, logic ~60. Speed rated high (90) due to small 8B parameter size.
Z.ai: GLM Flash Latest$2.3280
85
87
95
95
1.0M
89HumanEval 94.2%, GPQA 85.7%, IFEval 88.0%, AIME 95.7%. Mapped directly to 0-100 scale. As a Flash tier model, speed is rated high (95), while coding/logic reflect strong but lightweight capabilities compared to flagship GLM-5.
LiquidAI: LFM2-24B-A2B$2.4060
71
44
97
55
128K
54Benchable.ai cites Coding 71%, Reasoning 50%, Instruction 38%, and Speed at 97th percentile. Mapped directly to 0-100 scale. As a lightweight 2B active MoE, it prioritizes speed over flagship-level logic and coding.
Qwen: Qwen3.7 Flash$2.5064
55
63
90
65
1.0M
61No exact SWE-bench or GPQA scores provided for Qwen 3.7 Flash. Inferred from Flash lightweight tier: Speed mapped high (90), Coding/Logic mapped lower (55). Multimodal supported natively.
Qwen: Qwen-Turbo$2.6057
50
50
50
50
131K
50
Inference.net: Schematron V2 Turbo$2.7043
20
35
95
10
128K
25No standard benchmarks provided. As a specialized 3B Turbo HTML-to-JSON extraction model, coding, logic, and math are inferred low, while instruction (schema adherence) is moderate. Speed is rated very high (95) due to its lightweight architecture.
Meta: Llama 3.1 8B Instruct$2.8051
45
48
90
25
131K
41MMLU 66.7%, GPQA 8.72%, MATH 15.56%, IFEval 49.22%. As an 8B lightweight tier, scores map to low/moderate logic (45) and math (25). Speed is rated high (90) due to its small, efficient architecture.
Mistral: Mistral Small 3$2.8070
75
75
90
75
33K
75HumanEval 88.41% maps to coding 75. GPQA Diamond 45.96% maps to logic 70. As a 24B 'Small' tier model, it scores lower than flagships but achieves high speed (90).
Amazon: Nova Micro 1.0$2.8066
68
63
95
75
128K
67HumanEval 81.1% (Coding 68), GPQA 40% (Logic 45), IFEval 87.2% (Instruction 80), GSM8K 92.3% (Math 75). As a 'Micro' tier model, speed is rated very high (95) while coding and logic reflect its lightweight, text-only nature.
inclusionAI: Ling 3.0 Flash Sante$2.9166
65
68
90
75
262K
69SWE-bench Pro at 56.6% and SWE-bench Vals at 65.2% map to 65 Coding. As a Flash-tier MoE (5.1B active), it scores lower on Logic (65) but excels in Speed (90). Reasoning tokens are supported.
inclusionAI: Ling 3.0 Flash Fin$2.9157
35
60
85
55
262K
52Used Ling 2.6 Flash proxy scores: GPQA 59.3% (Logic 59) and Coding 25.3 (Coding 35). As a Flash-tier lightweight MoE, capability scores are kept lower than flagships, while Speed is rated high (85).
Cohere: Command R7B (12-2024)$3.0052
35
48
90
40
128K
43Evidence explicitly states no benchmark data is available, requiring a conservative rating. As a 7B lightweight model, capabilities are estimated lower than flagships, while speed is rated high (90) due to its small, fast architecture.
IBM: Granite 4.1 8B$3.0064
68
64
90
65
131K
65HumanEval >89.7% (coding ~68), MMLU >66% (logic ~58), IFEval >74.8% (instruction ~70) based on 3.3 baseline. Lightweight 8B tier adjustments applied for high speed and scaled capabilities.
Meta: Llama 3.2 1B Instruct$3.0945
25
35
95
30
60K
31MMLU 49.3%, GSM8K 44.4%, MATH 30.6%. As a 1B lightweight model, Logic (35) and Math (30) reflect sub-50% benchmarks. Coding (25) based on 0.6 index. Speed (95) is maximized for this ultra-small tier.
Inception: Mercury 2.5$3.1071
75
75
100
88
260K
78Based on GPQA (73.6-77%), HumanEval (85%), and IFBench (71.3%), mapped to mid-tier 70-80s. Speed is exceptional at ~1,107 tps (100). As a speed-optimized dLLM, it trades flagship coding for extreme throughput and tunable reasoning.
Inception: Mercury 2.5 Preview$3.1052
45
45
100
45
260K
45Fast-dLLM v2 averages 43.5 across HumanEval, GPQA, and MMLU, mapping to ~45 for coding/logic. As a 1B-scale lightweight model, it prioritizes speed (>1,000 tok/s, scoring 100) over flagship-level reasoning.
DeepSeek: DeepSeek V4 Flash 0731$3.1175
88
87
92
82
1.0M
86SWE-bench Verified (79.0%) and GPQA Diamond (88.1%) map to 88 coding and 89 logic. As a Flash tier (13B active), speed is rated high (92), though native reasoning modes boost its benchmark scores significantly.
NVIDIA: Nemotron Nano 9B V2$3.2066
70
63
90
85
131K
70Evidence shows GPQA at 64.0% (Logic ~65) and LiveCodeBench at 72.4% (Coding ~70). MATH-500 is 97.8% (Math ~85). As a 9B lightweight reasoning model, it achieves high speed (115 tok/s, Speed ~90) but lacks multimodal support.
Arcee AI: Trinity Mini$3.3075
82
89
90
88
131K
87GPQA Diamond at 92.1% maps to 92 Logic. AIME 2025 at 58.6% maps to 88 Math. LM Market Cap coding score of 82 maps to 82 Coding. As a 'Mini' tier, speed is rated high (90).
DeepSeek: DeepSeek V4 Flash Latest$3.4575
85
88
92
88
1.0M
87SWE-bench Verified 79.0% maps to 85 coding. GPQA Diamond 88.1 maps to 90 logic. As a Flash tier model, speed is rated high (92), though its native reasoning modes elevate logic scores near flagship levels.
Google: Gemma 3 12B$3.5067
70
70
88
85
131K
74HumanEval 85.4% (Coding ~70), GPQA 40.9% (Logic ~55), IFEval 88.9% (Instruction ~85), MATH 83.8% (Math ~85). As a 12B lightweight tier, scores reflect strong math/instruction but moderate logic/coding compared to flagships.
Tencent: Hy-MT2-1.8B$3.5339
20
25
85
20
8K
23Evidence lacks specific benchmark percentages for HY-MT2-1.8B, noting only 18 tokens/sec on M4. As a 1.8B lightweight translation model, coding, logic, and math are inferred low (~20-30), while speed is rated high (85) for its compact size.
Poolside: Laguna XS 2.1$3.6068
88
73
92
70
262K
76SWE-bench Verified at 68.2% maps to 88 Coding. As an XS (33B/3B active) lightweight tier, Speed is high (92). Logic (70) and Math (70) are inferred due to lack of GPQA/MATH evidence.

Need a shareable artifact?

Get a print-ready PDF of your results and a CSV spreadsheet. Tap the button, then enter your work email. We use it to build your files and start the download—and to email you a copy if the site owner enabled that.

AI ROI Leaderboard & Discovery by LeadsCalc

Detailed analysis

PDF Breakdown

Receive a comprehensive native vector PDF of this leaderboard: your workload, filters, top rankings, and a table snapshot (sorted: Cheapest (est. monthly)).

Instant setup
No CC required

By submitting, you agree to our Privacy Policy and Terms.

Agency accelerator

Whitelabel Est. monthly Leaderboard
for your site

Embed the interactive cheapest (est. monthly) view on your own domain — whitelabel branding, lead capture, and the same workload sliders your prospects already use on LeadsCalc.

1-Click CRM sync
Custom branding
Branded reports
Lead analytics

Free to start

$0/mo*
GET STARTED

NO CREDIT CARD REQUIRED

How it works

Methodology: How we rank Cost-Optimized LLMs

Transparent, benchmark-driven rankings—same craft as our single-model deep dives.

How batch and prompt caching affect “cheapest” rankings

Rankings are based on the blended cost of input and output tokens for a comparable monthly workload. Where our database verifies eligibility, we factor in typical 50% batch API discounts and prompt-caching savings (caching is applied in estimates only when our pipeline has confirmed supports_caching for that model). Figures are indicative for buyers comparing vendors transparently—not a quote from any single provider.

Battle Arena

Compare up to four LLMs side by side

Tick up to four models in the leaderboard table, then open Battle Arena for API pricing, benchmarks, and workload math in one view—perfect when you are shortlisting vendors for a pilot in the US, Canada, or Australia.

Prefer a head start? Jump into high-intent comparisons people search for every day—same interactive calculator, zero signup.

Open Battle ArenaUp to 4 models · Live estimates
Signals & spend

Value analysis

Benchmarks vs. estimated API cost—read the story your CFO cares about.

When the cheapest model is not the best business choice

Extreme savings can hide weaker reliability or context limits. Use this tab to find a cost floor, then cross-check coding or reasoning tabs for quality gates. Enterprise buyers in Australia and Canada often layer vendor DPAs and region routing on top of list economics; US teams frequently validate support SLAs before migrating traffic.

Production deployment

High-Volume Data Processing

How teams in the US, Canada, and Australia deploy these models in production.

Bulk extraction, sentiment analysis, and content moderation at scale

When processing millions of rows of unstructured data, token economics dictate feasibility. Data engineering teams in the US and Canada use these cost-optimized models for bulk entity extraction, continuous social media sentiment analysis, and automated content moderation, where 'good enough' accuracy at 1/100th the price of a frontier model creates massive business value.

Architecture

Maximizing API Cost Efficiency

Strategies to reduce monthly API spend without sacrificing capability.

100% batch API utilization and aggressive prompt caching

To achieve the absolute lowest cost per million tokens, modern architectures combine two features: Batch APIs (which typically offer a 50% discount for 24-hour turnaround) and Prompt Caching (which discounts large, static system prompts by up to 90%). Filter this leaderboard by 'Batch pricing' to identify vendors that support these aggressive cost-reduction features.

Embed-ready

Need this live Cost-Optimized data on your website?

Join 500+ agencies in the US and Australia using LeadsCalc to capture high-intent leads. Embed this interactive Cost-Optimized leaderboard on your site in about a minute—Canadian teams use the same flows for CAD-priced proposals and compliance-friendly landing pages.

Customize & Embed this ToolWhite-label · No code required
United StatesCanadaAustralia
Live preview

Your visitors compare Cost-Optimized models without leaving your domain.

Support & clarity

Frequently Asked Questions

Focused on teams in the United States, Canada, and Australia.

Start with realistic token and request volumes, enable batch pricing only when your integration supports it, and turn on cached pricing when your model and provider document cache tiers. Agencies across the US, Canada, and Australia use this table to shortlist models before negotiating enterprise commits.