Tools · Computational Biology

Evo 2

Genome modeling and design across all domains of life with the Evo 2 DNA language model

On this page

Prices, workflows, and method papers Open in the app

Run it from the API

Submit with Submit a job and the job_type below. Price it first with Estimate job reservation cost: submitting reserves that amount from your wallet, and the charge settles at the actual runtime.

Evo 2 `evo2`

Job type: evo2
Hardware: l4a10 (default)l40sa100-40gba100-80gbh100h200b200
Typical runtime: 1 h on Nvidia A10

Payload

Payload fields
Field	Type	Description
`task`required	`string`	One of: `"generate", "embeddings", "score"`
`model_name`	`string`	Default: `"evo2_7b"`One of: `"evo2_40b", "evo2_7b", "evo2_20b", "evo2_40b_base", "evo2_7b_base", "evo2_1b_base", "evo2_7b_262k", "evo2_7b_microviridae"`
`input_mode`	`string`	Default: `"sequence"`One of: `"sequence", "fasta", "pdb", "pdb_id", "csv", "excel"`
`sequences[]`	`string[]`	Limits: `min items 1`
`fasta_content`	`string`	Limits: `min length 1`
`pdb_content`	`string`	Limits: `min length 1`
`pdb_id`	`string`	Limits: `min length 1`
`pdb_chain_ids[]`	`string[]`
`csv_content`	`string`	Limits: `min length 1`
`excel_content`	`string`	Limits: `min length 1`
`sequence_column`	`string`	Limits: `min length 1`
`expected_sequence_count`	`integer`	Optional billing/runtime reservation hint. May raise the inferred sequence count for structure/Excel inputs, but cannot understate counts inferred from sequence/FASTA/CSV inputs. Limits: `≥ 1`
`generate_args`	`object`
generate_args.`n_tokens`	`integer`	Limits: `≥ 1`
generate_args.`n_sequences`	`integer`	Limits: `≥ 1`
generate_args.`temperature`	`number`	Limits: `≥ 0`
generate_args.`top_k`	`integer`	Limits: `≥ 1`
generate_args.`top_p`	`number`	Limits: `≥ 0`
generate_args.`batched`	`boolean`
generate_args.`cached_generation`	`boolean`
generate_args.`verbose`	`integer`	Limits: `≥ 0`
generate_args.`force_prompt_threshold`	`integer \| null`	Limits: `≥ 1`
generate_args.`use_kernels`	`boolean`
`embedding_args`	`object`
embedding_args.`layer_names[]`required	`string[]`	Limits: `min items 1`
`score_args`	`object`
score_args.`batch_size`	`integer`	Limits: `≥ 1`
score_args.`prepend_bos`	`boolean`
score_args.`reduce_method`	`string`	One of: `"mean", "sum"`
score_args.`average_reverse_complement`	`boolean`

Example

from cognichem_client import CogniChem

client = CogniChem.from_env()  # reads COGNICHEM_API_KEY
payload = {
    "input_mode": "sequence",
    "sequences": ["ACGT"],
    "task": "score",
    "model_name": "evo2_7b",
}

estimate = client.jobs.estimate(job_type="evo2", payload=payload, resource="a10")
print(f"Reserves ${estimate.cost:.2f}")

job = client.jobs.submit(
    job_name="my-evo2-run",
    job_type="evo2",
    payload=payload,
    resource="a10",
)
status = client.jobs.wait(job.process_id)
if status.status == "completed":
    client.jobs.result(job.process_id, save_path=".")

curl -X POST "https://api.cognichem.com/api/v1/jobs/submit" \
  -H "X-Api-Key: $COGNICHEM_API_KEY" \
  -H "Idempotency-Key: $(uuidgen)" \
  -H "Content-Type: application/json" \
  -d '{
  "job_name": "my-evo2-run",
  "job_type": "evo2",
  "payload": {
    "input_mode": "sequence",
    "sequences": [
      "ACGT"
    ],
    "task": "score",
    "model_name": "evo2_7b"
  },
  "resource": "a10"
}'

{
  "input_mode": "sequence",
  "sequences": [
    "ACGT"
  ],
  "task": "score",
  "model_name": "evo2_7b"
}

Sample data from the job catalog; long values are shortened here. Each job_name must be unique among your jobs.

Workflow inputs

None

Workflow outputs

ArchiveZIP
TableCSV
SequencesFASTA

Run it from the API

Evo 2 evo2

Evo 2 `evo2`