Cost of Screening Resumes with an LLM at Scale
Each resume is a few thousand tokens in and a short score and rationale out, so cost scales linearly with applicant volume.
Typical monthly workload
- Input tokens
- 15,000,000
- Output tokens
- 1,500,000
- Requests / day
- 2,500
Illustrative estimate. Adjust the sliders below to match your own volume.
Recommended setup
Extract structured fields with a small model, then compare against the job rubric. Keep a human in the loop for final decisions.
Monthly cost by model for this workload
| Model | Input / 1M | Output / 1M | Est. monthly cost |
|---|---|---|---|
| Gemini 1.5 Flash | $0.075 | $0.30 | $1.57 |
| DeepSeek V3 / V2.5 | $0.140 | $0.28 | $2.52 |
| GPT-4o Mini | $0.150 | $0.60 | $3.15 |
| Llama 3.3 70B (Groq/Together) | $0.590 | $0.79 | $10.04 |
| Claude 3.5 Haiku | $0.800 | $4.00 | $18.00 |
| o1-mini | $1.100 | $4.40 | $23.10 |
| Gemini 1.5 Pro | $1.250 | $5.00 | $26.25 |
| Mistral Large 2 | $2.000 | $6.00 | $39.00 |
| GPT-4o (Omni) | $2.500 | $10.00 | $52.50 |
| Claude 3.5 Sonnet | $3.000 | $15.00 | $67.50 |
| o1 (Reasoning) | $15.000 | $60.00 | $315.00 |
Computed from list prices and the workload above (15,000,000 input and 1,500,000 output tokens per month), with no caching or batch discounts. Real bills vary.
How to cut this cost
- Parse once, score many
Convert each resume to compact JSON once, then reuse it across multiple openings instead of reprocessing the raw file.
- Audit for bias
Screen on skills and evidence, not names or demographics, and review model decisions regularly.
Simulate your own spend
Quick Workload Presets
Pick a ready-made scenario or customize your token volume below.
Token & Usage Parameters
By switching or routing background tasks from GPT-4o to Gemini 1.5 Flash, your team saves up to $2,270 annually.
Lowest cost per token, high-frequency tasks, real-time audio/video
State-of-the-art open economics, high-performance general intelligence
High-volume text processing, fast chatbots, classification
Open-weights privacy, self-hosting or ultra-cheap hosted inferencing
Rapid code completion, customer service, structured data extraction
Fast STEM reasoning, competitive programming at lower cost
Massive context (2M tokens), full video/audio analysis, large codebases
Multilingual tasks, European compliance, enterprise code generation
Complex multimodal reasoning, coding, and production apps
Software engineering, nuanced writing, complex agentic workflows
Advanced mathematics, algorithmic coding, PhD-level scientific logic
Frequently asked questions
Is automated resume screening risky?
It can be. Hiring tools may be regulated in some jurisdictions and can encode bias, so keep human review and check local rules before relying on scores.