These are relative terms. Higher side of the spectra are some antibody antigen reactions and glycoprotein harmone and receptor-interactions can have affinities as high as nano molar to picomolar. Lower end would be micromolar in many of the metabolic substrates like house keeping enzymes where substrates are locally enriched.
Yaseer, it may help you to actually find and compare Kd values of strong versus weak interactions as suggested by Dr. Vijayalaxmi. Are you doing docking studies to find if your ligand binds a target with high affinity? Your answer will help us understand the problem and give a better answer!
In Biochemistry or Pharmacology, we consider a protein-ligand complex having high affinity if the Kd is below 100 nM (for antibody-antigen complex below 10 nM), medium affinity in the range 100 nM - 10 uM, and low affinity if the Kd is above 10 uM.
I agree with many of the answers above. We work with aptamers-target complexes and generally in these systems with protein targets high affinity would be nanomolar and below, medium affinity would be in the range of 100 nM to 1 uM and low affinity would be above 1 uM. However, in aptamers for small molecules rather than for proteins we would shift the range slightly since their affinity is usually weaker.
I have seen an arbitrary but useful assignment of Kd as 'weak' when it is 1 mM to 1 M, 'marginal' when Kd = 1 uM to 1 mM, 'strong' when Kd = 1 nM to 1 uM, and 'very (extremely) strong/tight' when the Kd = 1 pM to 1 nM. It is of course arbitrary and the above mentioned affinity nomenclatures are also correct. Small compounds, e.g. sulfate anion may have a Kd to a protein = 100 uM to 100 mM, regular biochemicals such as ATP will have Kd = 10 uM to 10 mM, many drugs will have Kd = 100 pM to 10 nM, many antibodies will have Kd = 10 pM to 1 nM, but there are many examples out of this range.