Changelog

What shipped in Talindrew, and when. Each entry is dated by the commit that landed it; nothing here is planned or in progress.

Last updated 26 Oct 2026. RSS feed

1. 26 Oct 2026Pricing

   Plus plan and a weekly free allowance for AI apps

   A new Plus plan at $9 a month or $99 a year, with a 7-day free trial, makes tool calls from Claude, ChatGPT, Codex, Cursor and other MCP clients unlimited. The Free plan keeps the whole cloning suite and lab notebook and now includes 25 MCP tool calls a week, reset every Monday at 00:00 UTC.

   • Plus adds no compute: protein structure prediction, FoldRx drug discovery runs and the AI assistant still start at Pro.
   • Nothing is charged during the trial, and Plus can be cancelled at any time.

   Pricing →

2. 25 Oct 2026MCP

   MCP: the cloning workspace, the lab notebook and sharing, from chat

   Thirty-eight more tools bring the MCP server to 90. An AI app can now build and edit a cloning workspace step by step, write lab-notebook entries with maps, primer tables, gels and cloning blocks in them, export entries as PDF, Word or Markdown, and share a sequence, workspace or entry with a colleague or by link.

   • The /mcp page gained two short films of Talindrew used from Claude and Codex, each with its own watch page, and a three-step connect guide.

   MCP server →

3. 24 Oct 2026MCP

   Setup pages for 23 MCP clients

   Each supported AI app has its own page with the exact steps, a configuration snippet or one-click install link where the client has one, a setup prompt to paste into an agent, and how to disconnect. Claude, Claude Code, Codex and ChatGPT were tested end to end against the production server.

   • ChatGPT custom connectors now complete sign-in: the server accepts the private_key_jwt client authentication ChatGPT declares.

   MCP clients →

4. 23 Oct 2026MCP

   Talindrew as an MCP server for Claude, ChatGPT and Codex

   A remote MCP server at https://mcp.talindrew.com/mcp (Streamable HTTP, sign-in with Google over OAuth) gives AI apps Talindrew's sequence tools: plasmid maps and restriction digests, primer design and Tm, CRISPR guides, alignment, translation and ORFs, gene, plasmid and literature lookup, and read and write access to your own sequences and notebook.

   • Inside Claude, results render as Talindrew panels: a primer's Tm shows the same four methods the app computes.
   • Results link into the app with their inputs filled in, so a calculation started in chat continues in the full editor.

   MCP server →

5. 2 Oct 2026Drug discovery

   Drug discovery hub, ESMFold2 and bring-your-own models

   FoldRx, the target-to-ranked-compounds pipeline, gained ESMFold2 (the single-sequence Fast variant) beside ESMFold and Boltz-2, and accepts your own model endpoint at the structure step, including AlphaFold 3 run on your own hardware under your licence.

   • A drug discovery hub with five guides: custom models, AlphaFold 3 commercial use, ESMFold2 vs AlphaFold 3 vs Boltz-2, structure-based virtual screening, and P2Rank vs fpocket pocket prediction.
   • Comparison pages against Neurosnap and Tamarind Bio, and a film of the pipeline in the AI workspace.

   Drug discovery →

6. 30 Sep 2026Pricing

   Free, Pro and Max plans

   The cloning suite and lab notebook stay free. Pro ($24 a month, $15 for academics) adds 50 protein structure predictions, 10 drug discovery runs of up to 100 compounds and 300 AI assistant messages a month; Max ($99, $69 for academics) raises those to 300 predictions, 50 runs of up to 500 compounds and 2,000 messages, and adds hosted models such as Boltz-2 and DiffDock.

   • Allowances reset on your subscription's own billing day, and a run that fails on Talindrew's side is not counted.

   Pricing →

7. 25 Sep 2026Site

   Video watch pages and animated explainers

   The product films have their own watch pages with chapters and transcripts, and calculators and guides gained figures that move with their inputs: the rotor on RPM to RCF, and the Gibson and Golden Gate mechanisms on their guides.

   Videos →

8. 15 Sep 2026Calculators

   Lookup charts, molarity conversions and batch reverse complement

   The free calculators now answer the numbers people look up most often without typing anything: server-rendered charts sit under each widget, and two calculators gained the conversions searches kept asking for. The Gibson and Golden Gate guides now carry NEB's step-by-step reaction setup.

   • RPM to RCF: ×g to RPM and RPM to ×g charts at 6, 8.4, 10, 15 and 20 cm rotor radii.
   • Molarity: µg/ml or mg/ml to molarity and molarity to mg/ml modes, plus charts by molecular weight and grams per litre for common buffer components.
   • Dilution: a 1:N chart with stock and diluent volumes for 1 ml and 10 ml finals.
   • Cell seeding: cells per well, dish and flask for 12 vessels at four seeding densities.
   • Reverse complement: paste several sequences as FASTA and every record is converted at once.
   • Gibson Assembly and Golden Gate Assembly guides: amounts, ratios, volumes and incubation programmes from NEB's E2611 and E1601 protocols.

   Calculators →

9. 9 Sep 2026Cloning

   Cloning workspace: a port of OpenCloning

   A node-graph cloning workspace that is a full port of OpenCloning (app.opencloning.org), running OpenCloning's own pydna engine inside Talindrew's backend. It saves OpenCloning's cloning-strategy JSON, so a strategy moves between the two tools in either direction, and it sits beside the existing cloning wizards rather than replacing them.

   • Sixteen operations on one graph: restriction digest, PCR, oligonucleotide hybridisation, polymerase extension, reverse complement, annotation, Gibson, In-Fusion, In vivo and overlap-extension assembly, restriction and ligation (Golden Gate), ligation, homologous recombination, CRISPR-HDR, Gateway BP and LR, Cre/Lox and recombinases with custom sites.
   • Sequences come in by hand, from GenBank, FASTA, EMBL or SnapGene files (including a .dna file's embedded cloning history), from Addgene, iGEM, SEVA, Euroscarf, WeKwikGene, Benchling, SnapGene or Open DNA Collections, or as a genome region from NCBI.
   • Every sequence node draws a real circular or linear map with a shape toggle; eleven example strategies are bundled and the draft survives a refresh.
   • Kit templates from the OpenCloning submission repository (MoClo YTK included), a Golden Gate assembler that chains category slots and assembles every valid combination, and a syntax builder with graph validation.
   • Primers tab with primer3 thermodynamics (Tm, GC, homodimer, hairpin, heterodimer per PCR) and csv/tsv import and export, plus a primer designer for normal PCR, restriction and ligation, Gibson-family assembly, homologous recombination and CRISPR, Gateway BP and EBIC that previews the product and writes the primers onto the PCR step.
   • Sequencing verification: AB1, SCF and ZTR traces aligned server-side, coverage drawn on both maps with identity, a mismatch table and the chromatogram, and the traces carried inside the exported strategy zip. This deployment aligns traces with Biopython rather than the MAFFT and MARS binaries the upstream engine shells out to.
   • A reaction saves into a notebook entry as a cloning block that links back to the workspace, and a workspace shares by read-only link or with named people as viewer or editor.
   • In the viewer: create, edit, duplicate and delete features from the right-click menu, create a primer from a selection, and new Properties and Lineage tabs.

   Cloning workspace →

10. 6 Sep 2026Primers

    PCR primer design

    A primer designer over any selected region of a sequence, plus an Oligos tab that is the construct's own primer library. The engine runs client-side, so the ranked candidate list re-ranks as the constraints move.

    • Per-primer scoring on Tm, GC content, length, 3′ GC clamp, homopolymer runs, hairpins, self- and 3′-anchored dimers and binding-site count; pairs scored on ΔTm, cross-dimers and product size.
    • Every Tm is the same nearest-neighbour value the Melting Temperature calculator uses, so the two cannot disagree.
    • Binding-site search on both strands with amplicon prediction, including products that cross the origin of a circular template.
    • 5′ overhang builder: 13 restriction sites with guard bases, Kozak, T7 promoter, His₆ tag, Gibson/HiFi homology arms or custom text; annealing Tm and full-oligo Tm reported separately.
    • Primers persist with the sequence, import from FASTA or a pasted list, export as CSV or FASTA, draw on both maps and export as primer_bind features.

    Primer design tool →

11. 6 Sep 2026Alignment

    Multiple sequence alignment

    A new Alignment workspace aligns up to 200 DNA, RNA or protein sequences. The built-in aligner is progressive: k-mer distances, a UPGMA guide tree, then profile-to-profile alignment with affine gap penalties, and MAFFT, MUSCLE and Clustal Omega are used instead when they are installed on the server.

    • Consensus row, per-column conservation track, identity-matrix heat map and a UPGMA tree built from the finished alignment.
    • Per-base colours for DNA, Clustal-X class colours for protein shaded by conservation, a differences-only mode and a click-a-column detail panel.
    • Load sequences from open tabs, a saved construct, a FASTA paste or a file; export aligned FASTA, CLUSTAL or Newick.
    • Eight 1.5 kb sequences align in well under a second; results are cached, so re-running an unchanged set costs no request.

    Multiple sequence alignment tool →

12. 5 Sep 2026Notebook

    Sample project, live maps in entries, image resizing

    Every new user now finds a seeded sample project on their first visit. Sequence blocks inside a notebook entry draw the real circular and linear maps, in the editor and in the print and shared read-only views. Images resize by dragging and keep their size everywhere the entry is rendered.

13. 4 Sep 2026Notebook

    PubMed and Europe PMC citations

    Search PubMed and Europe PMC from the browser and cite a paper straight into a notebook entry. Print exports now match the on-screen sheet in both the PDF and the Word file.

14. 3 Sep 2026Sequence editor

    SnapGene-style history and print preview

    Talindrew reads the cloning history tree out of SnapGene .dna files and records its own operations in the same model, drawn as a History tab. Click selection on the maps matches SnapGene's behaviour. A print preview of any sequence downloads as PDF or Word.

    Talindrew vs SnapGene →

15. 2 Sep 2026Notebook

    Electronic lab notebook with eLabFTW sync

    A full electronic lab notebook shipped in twelve phases: rich-text and markdown entries with equations, code and images; Talindrew blocks that save a cloning result, map, gel or trace into an entry; typed fields, templates, protocols with checklist steps and structured tables; comments, review, witness and e-signatures with timestamps; resources and inventory with storage, containers, compounds and QR labels; equipment booking with iCal export; export to PDF, ZIP, JSON, CSV and the .eln archive; and a two-way sync with eLabFTW.

    Talindrew vs Benchling →

16. 1 Sep 2026Platform

    Job queue for heavy workloads, Google sign-in

    Cloning, folding and other heavy requests run through a Redis-backed job queue with a separate worker service, so the API stays responsive and cloning no longer depends on a Temporal server. Sign-in is Google only.

17. 31 Aug 2026Calculators

    Biomath calculator suite

    Twenty-plus free calculators that run entirely in the browser: nucleic-acid mass and mole conversions, oligo Tm by four methods, ligation ratios, OD260, dilutions, molarity, qPCR efficiency, doubling time, g-force, each with its formula, citation, worked example and the reference calculator it is verified against.

    All calculators →

18. 22 Aug 2026Site

    Comparison pages, guides and tool pages

    Honest comparison pages against Benchling, SnapGene, Geneious and ApE with rows Talindrew loses; technique guides for Gibson Assembly and Golden Gate; tool pages for the restriction digest simulator and oligo Tm calculator; llms.txt for answer engines.

    All features →